Datasets:

OpenScholar

/

OpenScholar-DataStore-V3

like 5

Follow

OpenScholar 23

medications taken increased, the likelihood of having poor treatment outcome was reduced by 21% (AOR = 0.79, 95%CI; 0.73, 0.87). Patients' lifestyles played a great role in affecting poor treatment outcome. As compared to their counterparts, the odds of having poor treatment outcome was 1.29 (AOR = 1.29, 95%CI; 1.02, 1.65) and 1.72 (AOR = 1.72, 95% CI;1.46, 2.02) higher for patients having (Table 2). Discussion The impact of the COVID-19 pandemic on the healthcare system and patients' care has been sustained as the pandemic continues. Even though adherence to measures has been variable and inconsistent, Ethiopian governments have been forced to adopt legal measures to contain the spread of COVID-19 infection, including short-term complete lockdown, social distancing, prohibition of social gatherings, and school closure. The pandemic has had a significant impact on high-risk groups such as patients with chronic medical conditions such as hypertension and diabetes, either directly or indirectly [12][13][14]. The purpose of this study was to investigate the effect of COVID-19 on the magnitudes of poor treatment control among ambulatory Diabetic and/or Hypertensive patients and its associated factor using a generalized linear mixed model. During the COVID-19 pandemic, the magnitudes of poor treatment increased significantly. This shift was most noticeable in the second and third three months following the first COVID-19 case detection in Ethiopia. This could be due to disruptions in regular care caused by restrictions on essential health service visits, which forced them to stay at home, as well as PLOS ONE Effect of COVID-19 on poor treatment control among ambulatory

248916414

Hypertensive and/or Diabetic patients limiting physical contact with health professionals. When compared to the pre-pandemic period, the magnitudes of individuals who had missed their appointments increased more than twice during the pandemic period. As a result, patients who do not have strict follow-up and miss their appointments are more likely to have poor treatment control. In multivariable binary logistic mixed model, several factors such as marital status, duration of follow up, presence of complication, types of medication, frequency of medication used per day, numbers of medication, behavioural factors such as hazardous alcohol use and sedentary lifestyle, and missing of appointment were significantly associated with poor treatment control among ambulatory Diabetic and/or Hypertensive patients. According to the findings of this study, married participants had a lower chance of having poor treatment than unmarried participants. This finding, however, was consistent with previous studies conducted in Pakistan and China, which found that being married reduces the likelihood of poor treatment control [15,16]. The possible reason for this finding could be that married individuals might get support from their partners which positively affects the adherence to control measures for their underlying conditions [17]. In agreement with the previous studies [18,19], this study also strikes that duration followup was the important and significant factor for poor treatment control among ambulatory Diabetic and/or Hypertensive patients. This could be justified by as the duration of follow-up increases; the chance to develop disease-related complications will be higher which results in poor disease control. Besides, having a longer duration of follow-up might compromise the patient's

248916414

beliefs about the effectiveness of medication and control measures [20]. Therefore, it is better to screen for disease-related complications to achieve good disease control. In terms of medications, those who take injectable medications had a higher risk of poor treatment control than those who took PO medications. However, the types of medications and the frequency with which they were taken per day were negatively associated with poor treatment control. Patients taking one (daily), two (BID), three (TID), and four (QID) times per day had a lower risk of poor treatment control than those taking five or more times per day. Furthermore, a patient who took more types of medication had a lower chance of having poor treatment control. Previous studies supported this evidence [21,22]. This could be explained by that dosage and regimen for administration of medication had a paramount effect on medication adherence which is vital for controlling disease progressions and preventing complications [23], there is also a fact that patients taking injections may find it difficult to adjust treatment without the immediate support of health care providers [5,11,24]. Besides, taking multiple medications had a synergistic effect for treating the disease and preventing its complications [25]. Hazardous alcohol use and sedentary behaviour were also the significant factors of poor treatment control among ambulatory Diabetic and/or Hypertensive patients. It is in agreement with previous studies [26][27][28]. Because there is no safe alcohol range for chronic disease patients, and physical exercise is an important component of lifestyle modification, lifestyle modification plays a significant role in chronic disease

248916414

management, with the ultimate goal of preventing disease progression and related complications in instances where a complete cure cannot be achieved [29,30]. Missed appointment during the pandemic period was significantly and strongly associated with poor treatment control with the likelihood of having poor treatment control for patients who have missed their appointments were two times higher as compared to their counterparts. This finding was supported by previous studies [31,32]. This could be justified by the fact that missed appointment is one dimension of adherence where strict adherence to medication and appointment is required for chronic disease management [33]. The current study used internationally and/or locally validated tools for measuring physical activity and hazardous alcohol use, and data were collected by trained and experienced nurses and medical doctors under close and supportive supervision. The respondents were also informed about the importance of the study and the confidentiality of personal data to gain the trust of respondents and minimize the non response rate. But this study was not free of limitations. The study includes two medical conditions, thus the factors for poor treatment control might be different for each disease entity attention should be given while interpreting the findings of the study. Since the study was facility-based there might be a risk of social desirability bias. Moreover, there might be a risk of misclassification bias because the outcome variables were ascertained by the physician assessment. Conclusion COVID-19 pandemic has significantly affected the treatment control of ambulatory Diabetic and/ or Hypertensive patients. Being married, the frequency and kinds of

248916414

drugs taken per day were negatively associated with treatment control. Whereas hazardous alcohol use, sedentary lifestyle, longer duration of follow up, having a disease-related complication, patients taking injectable medication, per day, and missed appointments during the pandemic of COVID -19 were positively associated with poor treatment control of ambulatory Diabetic and/ or Hypertensive patients. Therefore, it is better to consider the risk factors of poor treatment control while designing and implementing policies and strategies for chronic disease control.

248916414

The social network: Neural control of sex differences in reproductive behaviors, motivation, and response to social isolation Social animal species present a vast repertoire of social interactions when encountering conspecifics. Reproduction-related behaviors, such as mating, parental care, and aggression, are some of the most rewarding types of social interactions and are also the most sexually dimorphic ones. This review focuses on rodent species and summarizes recent advances in neuroscience research that link sexually dimorphic reproductive behaviors to sexual dimorphism in their underlying neuronal circuits. Specifically, we present a few possible mechanisms governing sexually-dimorphic behaviors, by hypothalamic and reward-related brain regions. Sex differences in the neural response to social isolation in adulthood are also discussed, as well as future directions for comparative studies with naturally solitary species. Introduction Rodent species form various social organizations, ranging from solitary to dyads and families, to groups of tens or even hundreds of individuals. Living in a group bears many advantages that increase evolutionary fitness, such as mutual protection, assistance in resource acquisition, and more mating opportunities [1,2]. However, it also carries some costs, such as having to share valuable resources and the need to fight for one's social ranking [3]. Above all costs, social species endure life-threatening stress in the case of forced social isolation [4,5]. Despite the heavy costs, all social individuals show strong motivation to seek for, approach, and physically interact with social stimuli [6], yet substantial sex-based differences in these behaviors have been reported [2,7e 11]. Despite major efforts and recent advances in pinpointing the key neural

233446934

mechanisms underlying social behaviors in rodent models, many open fundamental questions still remain. Particularly, we have limited understanding regarding the extent of sex differences in the mechanisms governing social behaviors: Are sexually dimorphic social behaviors controlled by dimorphic brain structures, circuits, or molecularly defined neuronal populations? Do specific social signals have a different incentive value for males and females, governing sex-specific behavioral responses? In this review, we discuss the recent literature regarding behavioral sex differences in social displays and social reward in reproductive behaviors, specifically parental care mating, and aggression, focusing on rodents. We then present an emerging neural and molecular circuit logic underlying these sexually dimorphic social behaviors, which includes several hypothalamic and reward-related brain regions that are anatomically and functionally interconnected. This is followed by a discussion of the sexually dimorphic effects of social isolation in adulthood on social behaviors and their governing circuitries. Finally, we survey the fascinating diversity of social organizations in wild rodent species, which are barely used in neuroscience research but can provide unique and novel insights. evolutionary forces drive clear distinctions between males and females in their sexual motivation [8]. In most mammals, males are driven to inseminate as many females as possible, which is why their sexual motivation remains constantly high, while females must choose the best mate available, due to their enormous investment in breeding [8]. Indeed, studies that employed operant or classical conditioning paradigms have repeatedly shown that male rodents are highly rewarded by various sexual stimuli [17e21], while females usually require specific contexts and timing

233446934

for sexual interactions to be rewarding [8,22e24]. As to parental care, in most mammalian species, it is considered a female-typical behavior, although it can also be executed by males [25,26]. Substantial sex differences, in terms of both quantity and quality, are found in all components of parental behavior (e.g. nest building, pup retrieval, licking, and grooming) [25,27,28]. In laboratory mice, for example, sexually naïve males often attack and kill unfamiliar pups [25,29], but cease attacking and become paternal for a short period after mating with a female [30e32]. In contrast, sexually naïve laboratory female mice show spontaneous parental care toward unfamiliar pups [33,34]. Motherepup interactions are highly rewarding. For example, postpartum mice will exert efforts to cross a barrier in order to reach their pups [35]. Mother rats will compulsively lever press for pups [36], and such pup rewards were even more salient for postpartum females than an artificial drug reward [37,38] or natural rewards, such as food [39,40]. Aggression is usually considered a male-typical behavior in common laboratory mice, as males present robust aggression when competing over territory or potential breeding partners [8]. Laboratory females are usually aggressive toward unfamiliar individuals only during lactation, defending their offspring (i.e. maternal aggression) [41]. However, sexually naïve females of either wild or some outbred lab mouse strains can also present robust aggression while establishing their own territory and their social ranking in the group [42e44]. Nevertheless, females are less likely to engage in physical assault, and their attack patterns are less robust compared to males [45]. Also, the

233446934

aggression of wild female rodents is claimed to be more influenced by environmental conditions, such as the sex composition of the group [46,47], and by their estrus phase [48,49]. For males, aggressive behavior can be as rewarding as sexual behavior [50e52] and may even show similarities to addictive behaviors [53]. Male mice will lever press for a subordinate intruder introduced into their cage to be attacked [53] and will prefer an aggression-associated chamber in a conditioned place preference paradigm [54,55]. Male mice will also cross an electrified grid [56] and exhort efforts [57,58] to reach and attack a subordinate. Whether aggression is also rewarding for females remains unresolved, since most domesticated female rodents typically do not present aggression outside the postpartum period [27,33,59e61]. Some insights might come from the Syrian hamster, where females showed a conditioned place preference for a chamber associated with aggressive encounters with males [62]. Interestingly, same-sex aggressive interactions were found to be rewarding in both sexes of the Syrian hamster [9,63,64] and prairie vole [65], though the rewarding value seems more robust in females than in males (see Figure 1). These remarkable sex differences in reward and motivated social behaviors suggest an underlying sexual dimorphism in the neural circuits or neuromodulatory systems regulating them. Notably, it is known that both sexes might retain the capacity to express the behavioral repertoire typical of the opposite sex. For example, it was shown that female mice can present male-typical mounting and courtship [66,67] and that sexually naïve males can perform female-typical pup retrieval [32]. Thus,

233446934

some of the neural circuits governing sexual dimorphism are shared between the sexes [67]. In the following sections, we will present the growing amount of data regarding cell-specific neuronal populations and their connecting neuronal network that govern sexually dimorphic reproductive behavior, and their crosstalk with the brain's reward system, focusing on mice and rats. Of important note, despite the abundant literature demonstrating quantitative and qualitative sex differences in behavior, our understandings of the neurobiology of many fundamental neuronal processes are interpreted from male-exclusive studies [59,68]. Moreover, as shown in Figure 1, nature provides a wide diversity of social and reproductive strategies even within the Rodentia order. Yet, modern neuroscientists have only recently accepted the importance of using ecologically relevant species in order to better understand the mechanisms underlying social behavior, in males and females. Are dimorphic behaviors driven by dimorphic circuits? Neuronal circuits underlying sex differences in social reward and behaviors A few brain regions, most of them located within the hypothalamus, have been identified as critical nodes in the rodent social network and have also been shown to be sexually dimorphic. Within each dimorphic brain region, only specific, molecularly distinct subsets of neurons have been found to be sexually dimorphic and required for one or a few specific displays of social behavior [27]. These brain regions, including the anteroventral periventricular nucleus (AVPV), medial preoptic nucleus (MPOA), ventromedial hypothalamus (VMH), bed nucleus of the stria terminalis (BNST), and medial amygdala (MeA), send and receive multiple projections from one another to form robust 'social networks.' These social

233446934

networks communicate with brain reward regions and play an essential role in the regulation of most, if not all, sexually dimorphic reproductive behaviors [69,70]. To what extent are these anatomically interconnected dimorphic brain regions involved in the regulation of distinct sex-specific social behaviors? And, in case the functions of the social brains of adult males and females are partially bivalent, to what extent can we experimentally disrupt the circuit to drive behavior that is typically presented by the opposite sex? Recent technological advances in neuroscience now allow anatomical and functional mapping of these circuits, providing new insights into these fundamental questions, described below for each of the key regions involved in sexually dimorphic behavior. The anteroventral periventricular nucleus Sexually dimorphic brain regions are usually larger in males than in females [71,72]. For example, the sexually dimorphic nucleus of the preoptic area [73], the posterodorsal MeA [74], and the lateral septum [75], are all larger in males. A notable exception is the AVPV of the hypothalamus, which is larger in volume, contains more cells, and sends more projections to multiple reproduction-related brain regions in females compared to males [25,34,71,72,76e79]. Importantly, it also expresses several sexually dimorphic molecularly defined neuronal populations, including the tyrosine hydroxylase (TH)-expressing population, which contains 3e4 times more neurons in females than in males [34,72]. Examining the role of sexually dimorphic TH þ AVPV neurons in mice reveals a sex-specific function: in females, these neurons regulate maternal behavior, and in males, they repress inter-male aggression [34]. Specifically, ablation of TH þ AVPV neurons in

233446934

females reduces maternal behavior and activation increases maternal Diversity of social strategies within the Rodentia order: the uncharted territories of social neuroscience. Within the Rodentia order, the social scale ranges from eusocial and social to solitary, even within the same subfamilies and between closely related species [42,219]. Interestingly, within these diverse social lifestyles, an additional layer of diversity exists with respect to sexual dimorphism in parental care and aggression [220]. Presented from top to bottom-eusocial rodents: naked mole-rat [219,221,222] [228,229], meadow vole [219,226]; and solitary: blind mole-rat [197,199,203,230], Patagonian tuco-tuco (picture courtesy of Prof. Annaliese Beery) [231,232]. The symbols \ (female) and _ (male) denote sexual dimorphism or sex similarities for each behavior within a given species. (*) In the eusocial naked mole-rat and Damaraland mole-rat, only the breeding indare aggressive.(#) In the Damaraland mole-rat, the level of aggression depends on the sex of the attacked individual. Toward other females, the breeding female is more aggressive compared to breeding males. In contrast, toward other males from outside the colony, the breeding males are more aggressive compared to the breeding females [225]. behavior, while in males, their ablation increases aggression, whereas optogenetic activation reduces aggression [34]. The activation of TH þ AVPV neurons also leads to increased oxytocin (OT) release from the paraventricular hypothalamic nucleus (PVN) in females but not in males, suggesting that this circuit governs maternal behavior through the regulation of neuropeptides [25,34]. Notably, although manipulation of TH þ AVPV neurons markedly altered sex-specific behaviors in both males and females, the behaviors displayed by manipulated

233446934

animals of both sexes remained within the boundaries of their sex-typical behaviors (i.e. manipulations of TH þ AVPV could not induce parental care in sexually naïve males or aggression in sexually naïve females) (see Figure 2a). The medial preoptic nucleus A large hypothalamic structure, the MPOA sends projections to multiple downstream brain regions and is both larger and contains more neurons in males than in females [35]. Notably, the MPOA is home to various heterogeneous, molecularly defined, neuronal clusters, including many sexually dimorphic populations, such as androgen receptor (AR)-expressing population and estrogen receptor alpha (ESR1)expressing population [80]. The MPOA has been shown to be strongly activated by sex-typical social behaviors, functioning as one of the main brain regions that control parental care At least two different subpopulations within the MPOA were shown to be required for the regulation of pupdirected behavior. The first is the ESR1 þ population, which is highly sexually dimorphic in its distribution and projection patterns [85]. Suppression of ESR1 expression in the MPOA significantly reduced maternal behavior, but not maternal aggression, in female mice [86], whereas optogenetic activation of this population increased pup retrieval in females and in castrated males [84]. The second subpopulation is the galaninexpressing (Gal þ ) neurons, which showed increased activity during parental behavior in both female and male mice. Ablation of these cells reduced parental behavior in parenting females and males, whereas optogenetic activation suppressed pup-directed aggression in sexually naïve males and increased pup grooming in sexually naïve males and in fathers [32]. Notably, it was reported

233446934

that both of these neuronal subpopulations are also involved in the regulation of sexual behavior in both sexes [83]. Taken together, it appears that multiple molecularly defined subpopulations within the MPOA, and their segregated neuronal circuits, control the same sex-typical social behaviors. The ventromedial hypothalamus Over the past decade, the central role of the VMH, and specifically its ventrolateral part (VMHvl), in the initiation and execution of aggression has been well established in laboratory male mice [76,87e92]. In addition, it was shown that the activation of this region is necessary for promoting aggression seeking (i.e. reward) in male mice [89]. The role of VMHvl in female Another molecularly defined sexually dimorphic VMHvl subpopulation that controls sex-typical behaviors in both sexes is the progesterone receptor (PR)-expressing neurons. This subpopulation is required for the normal display of mating in both sexes and for fighting in males [76]. Ablation of PR þ VMHvl neurons led to a profound decrease in female sexual receptivity and in male sexual behavior and aggression. Moreover, it was recently shown that projections of PR þ VMHvl neurons into the AVPV nucleus change across the female mouse estrous cycle, with connectivity and function profoundly increasing during the estrus phase. This fluctuation in connectivity was found in adult females, but not in adult males, and was regulated by estrogen signaling in PR þ VMHvl neurons [94]. These findings highlight, once more, the critical role of intrinsic sex differences in the brain in setting the distinct behavioral repertoire displayed by each sex, as detailed above for the

233446934

AVPV. Furthermore, transcriptome analysis of VMHvl subpopulations provides evidence that different molecularly defined dimorphic populations in males and females may drive similar social behavior displays (i.e. fighting behaviors) [95] (Figure 2b). The bed nucleus of the stria terminalis The BNST is a critical component in the social behavioral network, interfacing with brain regions that are essential for social decision-making and reproductive behaviors [96e98]. The BNST is anatomically and functionally connected to many brain regions shown to regulate sexually dimorphic behaviors [69,96,99]. Recently, it was shown that ablation of the sexually dimorphic male-biased AVP þ BNST neuronal population reduces maleemale social investigation and increases scent marking, but does not affect aggression or other sexual behaviors of male mice. In females, the same ablation altered sexual behavior alone [100] ( Figure 2a). Several experiments on California mice (Peromyscus californicus) have demonstrated the key role of the BNST in the control of sex-specific behavioral responses to social defeat stress (for a review, see Ref. [99]). For example, it was shown that social defeat decreases the number of social approaches and increases OT þ BNST neuronal activity following social interaction, but only in females [101], while knockdown of OT in the BNST of females prevented the social-defeat-induced reduction in social approaches [102]. The role of OT þ BNST neurons in males remains to be elucidated. The medial amygdala The MeA is a nucleus within the amygdalar complex, which receives essential social information from the vomeronasal system and relays these pheromonal signals to the rest of the Sex differences in the

233446934

response to social isolation Social isolation induces severe stress in social species and might even be life-threatening [4,5,121e124]. In rodents, social isolation, during either adolescence or adulthood, was shown to dramatically disrupt behavior and brain function, in a sex-specific manner [125]. The effects of postweaning social isolation (i.e. adolescence), which are distinct from those of social isolation at adulthood [126,127], have been extensively reviewed [121,123,128] and will not be discussed here. In adult mice and rats of both sexes, even a brief period of social isolation can cause an aversive, 'loneliness-like' brain state [129], prompting animals to seek social interactions [129e133] and elevating the salience of social reward [121,131]. Social isolation in rodents also leads to many negative behavioral effects in both sexes, including increased territoriality and aggression [123,133e140], elevated anxiety-related behaviors [141], and depression-like symptoms [141e144]. In socially monogamous adult prairie voles, for example, both sexes display depressive-like and anxiety-related behaviors when separated from their bonded partner of the opposite sex [142,145e149] (Figure 1). Nevertheless, these effects seem to inflict females more than males [123, 143,150e152]. For example, in female mice, individual housing appears to increase plasma corticosterone and anxiety levels in the elevated plus maze assay [153] and in a modified open-field test [150], compared to females living in group housing. In contrast, individually housed males were not a affected, and their levels of anxiety and plasma corticosterone did not differ from group-housed controls [150,154e156]. Socially isolated adult female mice also displayed increased immobility in the forced-swim and tailsuspension tests, indicating depression-like symptoms [157],

233446934

while similar effects in males were induced only when isolation occurred throughout adolescence [158] or after prolonged isolation [127, 144,159,160]. In prairie voles, a 4-week isolation period reduces sucrose preference and increases corticosterone secretion in females, but not in males [143,145]. Social isolation also produces sexually dimorphic responses when it comes to sexually dimorphic reproductive behaviors. The social isolation of lactating female mice and rats reduced the duration of maternal care [161,162], while the isolation of male mice actually induced paternal behaviors toward unfamiliar pups and, in line, reduced their typical infanticide response [163,164]. Despite these profound sex differences in the behavioral output of social isolation in adulthood, we know very little about the neural mechanism underlying this dimorphism. In adult rats, prolonged isolation reduced spine density and the expression of synaptic proteins in the prefrontal cortex (PFC) of both males and females; in females, however, these parameters were also robustly influenced by the estrus state, with elevations during proestrus [144]. Similarly, social isolation reduced the expression of myelin transcripts in the PFC of both male and female mice [165]. Conversely, prolonged isolation increased the total expression of brain-derived neurotrophic factor (BDNF) and cAMP response elementbinding protein (CREB) in the cerebral cortex of female mice [166]. In the hippocampus, the social isolation of adult mice and rats led to reduced BDNF expression in both males [141,167e169] and females [167,170], and to increased neurogenesis in males [171] as well as females [172]. Notably, sex differences in the neural responses to social isolation were identified in the neuromodulator

233446934

systems of OT and arginine vasopressin (AVP). In socially isolated male mice, OT plasma levels were elevated [155], with IP administration of OT abolishing isolationinduced aggression [134], while in isolated female rats, the baseline OT levels measured in the CSF were similar to those of group-housed controls [135]. In socially monogamous prairie voles, prolonged isolation reduced oxytocin receptor (OTR) expression in the hypothalamus of both sexes [173], but elevated plasma OT levels [143,173] Figure 1), isolation reduced OTR levels in the dorsal raphe nucleus (DRN) and increased OTR levels in the anterior hypothalamus, but both effects were seen only in females [133]. The isolation also reduced the levels of the vasopressin V1a receptor in the BNST of both sexes, but in the DRN, V1Ra levels were reduced only in males [133]. Finally, a recent study on mice revealed a brain-wide signaling mechanism that mediates the effects of adult isolation on aggressive and anxiety-related behaviors [136]. The researchers noted a massive increase in the expression levels of the neuropeptide Tachykinin 2 (Tac2) throughout the brain of socially isolated mice of both sexes [136]. Further viral and pharmacological manipulations in males revealed that blocking Tac2's increase in the dorsomedial hypothalamus abolished social isolation-induced aggression, while blocking Tac2's increase in the central amygdala (CeA) abolished both acute and persistent stress responses [136]. Interestingly, manipulating Tac2 in the CeA produced sexopposite effects on fear learning in mice. While the administration of a Tac2 antagonist, or chemogenetic inhibition, impaired fear memory in males, it enhanced fear memory consolidation in females, and

233446934

both effects were mediated by sex hormones [177]. All in all, most of the studies involving the social isolation of adult rodents showed similarities in the effects on neural plasticity in both sexes. Thus, the robust sex differences in the behavioral responses to isolation might be attributed to sex differences in various neuropeptide systems, such as OT [125, 178,179] and AVP [125,180], or perhaps to fluctuations in the stress response throughout the estrus cycle [181]. Further studies are needed to unveil the underlying molecularly defined populations within the above-described brain regions that contribute to sex-differences in the behavioral effects of social isolation in adulthood. A special insight can be gained from observing the effects of social isolation on animal species maintaining a eusocial lifestyle. These species live in large communities, where typically only a few individuals bear offspring, while all the others share the burden of foraging food and caring for the young [42] (see Figure 1). In the eusocial naked mole-rat (Heterocephalus glaber) individuals removed from the colony displayed robust and stable increases of corticosterone levels for days and weeks [182,183] and increases in same-sex aggression among females [183]. Only two mammalian species are known to maintain a eusocial lifestyle, the naked mole-rat and its evolutionary relative, the Damaraland mole-rat (Fukomys damarensis) [42, 184], and in male and female breeders of both species the PVN is significantly larger [185,186]. Notably, in naked molerats, this effect is triggered by social isolation (i.e. removal from the colony) and not by breeding itself [187]. Also, in both species

233446934

breeders of both sexes display lower levels of hippocampal neurogenesis compared to subordinates [188,189]. Although further research is needed in males and females of eusocial species, the available literature suggests an intriguing hypothesis: that their level of sexual dimorphism is substantially reduced compared to other social species, both for processing of social stimuli, as well as in the response to social isolation (see Figure 1). Conclusions and future research suggestions The evidence that males and females are able to present most of the social behavioral repertoires typically presented by the opposite sex indicates that components in the underlying circuits driving sex-typical behaviors are shared between the sexes. The accumulating recent studies in the field indicate that within these shared neuronal components (social network), there are distinct molecularly defined neuronal populations that are anatomically and functionally dimorphic. These neuronal populations are those orchestrating the degree of and the manner by which specific sex-typical behaviors are displayed in each sex and each social condition. Sexually dimorphic behaviors can be governed by at least three possible neural mechanisms (see Figure 2) [13,27,67]: (1) A sexually dimorphic neuronal circuit, which promotes a distinct sex-specific behavior in each sex Despite the accumulating recent studies laying out the mechanisms underlying sex-typical social behavior, it seems that most of them were performed exclusively in males. Thus, in many cases, it is impossible to draw a clear conclusion whether the same network underlies same or sex-specific social behavior, or whether a sexually dimorphic brain region contributes to sex differences in social behavior. Moreover, classical

233446934

neuroscience tools were built for and applied almost exclusively on laboratory (i.e. inbred and domesticated) mice, which present an altered, artificially selected social behavior repertoire, compared to wild mice [59]. This is particularly evident in females, and specifically regarding reproductive behaviors such as aggression and prenatal care [25,33]. It remains to be determined whether the governing neuronal principles underlying sex-typical social behavior, discovered in inbred domesticated rodents, can also be applied to wild-derived rodents. It is repeatedly stated that scientists must expand the diversity of animal models beyond laboratory mice and rats, which are the subjects of more than 80% of current scientific research in animals [59,192,210e214]. The minority of neuroscience studies employing this approach have found unique phenotypes, alongside unique neural mechanisms underlying social behavior. For example, research on social ranking in the naked mole-rat showed higher oxytocin levels in subordinate individuals [215,216], while crossspecies comparative analysis revealed a role for oxytocin in the monogamous behavior of prairie voles [217], and for vasopressin in parental behavior of deer-mice (Peromyscus) [218]. In summary, further research with various animal models, including highly social or solitary mammalian species (see Box 1 and Figure 1), as well as research on the neural effects of social isolation in both sexes, in a comparable manner, is needed [5]. This should provide a more comprehensive understanding of the mechanisms through which behavior alterations and neuropathologies like depression are induced by social isolation, as experienced worldwide during the COVID-19 pandemic, or by perceived social isolation (i.e. loneliness). Conflict of interest statement Nothing declared.

233446934

References Papers of particular interest, published within the period of review, have been highlighted as: The authors nicely demonstrate that rats searching for a water reward in groups find the reward much faster than when attempting to do so alone. Notably, their findings indicate that females benefit more from group searching: the magnitude of improvement from single to group searching was higher in females, with female groups also performing significantly better than male groups. 3. Lee W, Yang E, Curley JP: Foraging dynamics are associated with social status and context in mouse social hierarchies. PeerJ 2018, 6, e5617. Box 1. Social isolation by nature: a-social behavior in wild rodent species. There are plenty of loners across the animal kingdom, including many mammalian species, ranging from primates to carnivores, to small rodents. The definition for solitary living species is that "the general activity, and particularly, the movements of different individuals about their habitat are not synchronized" [190], i.e. they usually live, sleep, and forage alone [191,192], with possible exceptions for mating and raising their young. The solitary lifestyle is highly common in subterranean taxa [193,194]. Interestingly, even within some families of subterranean rodents, social strategies range from solitary to eusocial [42,193,195], providing a unique opportunity to perform comparative studies and investigate the neural and evolutionary substrates driving the transitions across the 'social scale' (see Figure 1). Among these subterranean rodents, the blind mole-rat (BMR, Spalax ehrenbergi) exhibits one of the most solitary and aggressive life strategies, with relatively low levels of behavioral sex differences [196][197][198][199][200][201] (see Figure

233446934

1). Each individual excavates its own tunnel system to fit its body width and never leaves it unless forced to do so [196,201]. In the lab, introducing two adults of any sex (same sex or opposite sex) into the same cage will immediately lead to severe aggression, which will likely end in critical injuries and death [202]. Even prolonged exposure to non-direct social stimuli (pheromones or seismic signals) will induce severe chronic stress, leading to illness and, eventually, death [203]. During the rainy breeding season, it appears that the males socially communicate with potential females from a long distance and orient themselves to the female territory [204,205]. The males resume their solitary lifestyle immediately after successful copulation, while the females care for the pups alone [197,199,201,204]. Therefore, in order to ensure survival and reproductive success, BMRs must constantly be aware of the location and social status (e.g. sex, reproductive stage, and dominance) of their neighboring BMRs. Consequently, BMRs have developed a set of unique light-independent sensory modalities, perfectly adapted for social communication in their underground niche [206,207]. For example, for long-distance communication between conspecifics, BMRs use vibratory (seismic) signals produced by tapping their head against the roof of the tunnel [208,209]. Unfortunately, virtually nothing is known about the brain regions, neuronal circuits, and neuromodulators driving social reward and behaviors in solitary species. Many basic questions remain open, such as what are the neuronal and molecular processes that 'guard' solitary species from the harmful physiological damages of social isolation? What changes occur in their brain during the

233446934

breeding period, when solitary individuals need to physically interact with conspecifics? We believe that the BMR, as well as other a-social, solitary rodents, can provide a natural powerful model with which to study the adaptive neural principles governing social brain plasticity and social isolation stress. Moreover, solitary models can assist in studying sex differences in behavior and in the brain functions underlying social withdrawal and antisocial symptoms in psychiatric disorders (see Figure 1). The authors present an interesting hypothesis, where they compare the kinetics of social reward to that of drug rewards, describing the rewarding value of social interaction as a function of the 'dose' (i.e. duration of social interaction). They assert that, at first, increasing the dose heightens the value of the social reward. However, after reaching a certain peak, the value of the reward decreases as the dose increases. Thus, the value of social reward as a function of dose produces an "inverted U" curve, similar to drug rewards. Notably, the authors claim that this "inverted U" curve is different for males and females. 96-108. This review highlights three major biases present in modern social neuroscience: species bias (i.e. relying mostly on laboratory mice and rats as animal models), setup bias (i.e. using standardized, artificial experimental apparatuses), and sex bias (i.e. assessing only one sex, usually males). This study characterizes the behavioral phenotype of mutant female mice that fail to detect pheromone inputs through the vomeronasal organ, showing that these females display male-typical courtship and mounting behaviors toward an intruder, be it male or

233446934

female. These findings suggest that male and female mating circuits exist in both sexes, but are activated or repressed by chemosensing-directed circuitry. This study demonstrates the essential role of the VTA-NAc DA pathway in mediating pheromonal-evoked sexual reward in male mice. It also shows that NAc core D1R signaling is crucial for establishing conditioned place preference for female pheromones. 1535-1551. The authors employ fiber photometry to demonstrate that DA + VTA neurons are highly active during the initial phases of a female -female social interaction. They show that optogenetic activation of these neurons, as well as their projection to the NAc core, but not to the prefrontal cortex, modulates the duration of a social interaction.

233446934

Fine mapping of a linkage peak with integration of lipid traits identifies novel coronary artery disease genes on chromosome 5 Background Coronary artery disease (CAD), and one of its intermediate risk factors, dyslipidemia, possess a demonstrable genetic component, although the genetic architecture is incompletely defined. We previously reported a linkage peak on chromosome 5q31-33 for early-onset CAD where the strength of evidence for linkage was increased in families with higher mean low density lipoprotein-cholesterol (LDL-C). Therefore, we sought to fine-map the peak using association mapping of LDL-C as an intermediate disease-related trait to further define the etiology of this linkage peak. The study populations consisted of 1908 individuals from the CATHGEN biorepository of patients undergoing cardiac catheterization; 254 families (N = 827 individuals) from the GENECARD familial study of early-onset CAD; and 162 aorta samples harvested from deceased donors. Linkage disequilibrium-tagged SNPs were selected with an average of one SNP per 20 kb for 126.6-160.2 MB (region of highest linkage) and less dense spacing (one SNP per 50 kb) for the flanking regions (117.7-126.6 and 160.2-167.5 MB) and genotyped on all samples using a custom Illumina array. Association analysis of each SNP with LDL-C was performed using multivariable linear regression (CATHGEN) and the quantitative trait transmission disequilibrium test (QTDT; GENECARD). SNPs associated with the intermediate quantitative trait, LDL-C, were then assessed for association with CAD (i.e., a qualitative phenotype) using linkage and association in the presence of linkage (APL; GENECARD) and logistic regression (CATHGEN and aortas). Results We identified four genes with SNPs that showed the

7523047

strongest and most consistent associations with LDL-C and CAD: EBF1, PPP2R2B, SPOCK1, and PRELID2. The most significant results for association of SNPs with LDL-C were: EBF1, rs6865969, p = 0.01; PPP2R2B, rs2125443, p = 0.005; SPOCK1, rs17600115, p = 0.003; and PRELID2, rs10074645, p = 0.0002). The most significant results for CAD were EBF1, rs6865969, p = 0.007; PPP2R2B, rs7736604, p = 0.0003; SPOCK1, rs17170899, p = 0.004; and PRELID2, rs7713855, p = 0.003. Conclusion Using an intermediate disease-related quantitative trait of LDL-C we have identified four novel CAD genes, EBF1, PRELID2, SPOCK1, and PPP2R2B. These four genes should be further examined in future functional studies as candidate susceptibility loci for cardiovascular disease mediated through LDL-cholesterol pathways. Background Coronary artery disease (CAD) is the end result of accumulation of atheromatous plaques in the coronary arteries, leading to eventual impairment of cardiac blood flow and potentially devastating consequences of myocardial infarction (MI) or death. CAD is the leading cause of death in both the United States and worldwide, with over 500,000 deaths per year in the U.S. and over seven million worldwide (World Health Organization) [1]. Despite the development of pharmacologic therapies for prevention, the incidence of CAD is increasing, concomitant with the rising prevalence of risk factors such as obesity and diabetes (American Heart Association) [2]. CAD itself is clearly a heritable trait, with the role of genetic factors becoming increasingly apparent in earlyonset CAD [3][4][5]. However, the genetic architecture of CAD, as with many common diseases, is assumed to be complex and continues to be

7523047

poorly understood. Candidate gene studies have identified several loci for CAD, but with inconsistent results in validation cohorts. Recent genome wide association studies (GWAS) have consistently identified a locus on chromosome 9p21; however, this locus confers only modest risk of disease with effect sizes of 1.3-1.6 [6]. Thus, much of the genetic architecture underlying the heritability of CAD remains to be elucidated. There are many well-established risk factors for CAD that are partitioned between extrinsic (smoking, sedentary lifestyle, poor nutrition) and intrinsic (sex, age, lipid levels, hypertension) factors, each of which may have underlying genetic components, making it difficult to divide CAD risk into genetic and non-genetic factors. However, using these intermediate disease-related intrinsic factors as genetic traits may help to identify novel CAD genetic loci. We have previously reported a genome-wide linkage scan for early-onset CAD using the GENECARD familybased cohort, which identified nine genomic regions linked to CAD [7]. The 1q and 3q regions have been fine-mapped and the susceptibility genes identified (including FAM5C and KALIRN, respectively) [8,9]. The signal for 5q31 was present in the overall sample and was not unique to any one phenotypic subset. However, using ordered subset analysis (OSA) to dissect genetic heterogeneity and using lipid levels as quantitative traits, we found that the evidence for linkage on chromosome 5q was increased in families with higher mean total and low density lipoprotein-cholesterol (LDL-C) [10]. Rather than focusing on disease status alone as the trait of interest using one analytic technique, one can apply multiple methods to CAD and disease-related traits

7523047

within one genomic region, thereby exploring the solution space of the combined analyses and identifying overlapping results; such results act as an internal replication and increase the likelihood that the genetic variant is truly involved in the pathogenesis of CAD. Recently Williams and Haines argued that the replication standard is a strong indicator of a true genetic effect and possibly preferable to the p-value standard [11]. Thus, we report herein our work to fine-map the CAD susceptibility locus on chromosome 5q31, using association analyses of quantitative (LDL-C) and qualitative (CAD and atherosclerosis) traits, using the quantitative results to prioritize the results obtained from qualitative analyses. We conducted this study in several relatively large and independent CAD cohorts, including 1908 individuals from a cohort of patients undergoing cardiac catheterization (CATHGEN), 827 individuals from a family-based study of early-onset CAD (GENECARD) and 162 individuals from a repository of aortic tissue collected from deceased donors. Using this approach of analyses performed in parallel, we identified four genes on chromosome 5q31-33 (SPOCK1, PPP2R2B, PRELID2, and EBF1) as candidate susceptibility genes for CAD mediated through LDL-C. Study populations All subjects signed a current informed consent form and these studies were approved by the institutional review boards of each participating center. GENECARD family-based study of early-onset CAD Genetics of Early Onset Cardiovascular Disease (GENE-CARD) is a multicenter family-based linkage study of early-onset CAD using an affected sibling pair based approach; study methods have been described [7]. For GENECARD, early-onset CAD was defined as: MI or unstable angina, coronary angiography showing at least

7523047

50% stenosis in a major vessel, revascularization procedure as either percutaneous coronary intervention or coronary artery bypass graft, or a functional test showing reversible myocardial ischemia, occurring before the age of 51 in men and before the age of 56 in women. Of the 438 families included in the original linkage study [7], we selected 254 families, including 726 individuals (504 affected and 222 unaffected) for analysis of cardiovascular endpoints. These families were selected based on the availability of an unaffected family member to maximize power for association analyses (151 families). In addition, families identified from OSA [10] that contributed to the linkage peak on chromosome 5 were also included (103 families). For the analysis of LDL cholesterol traits, 827 individuals from these 254 families were used. LDL-C values were either extracted from medical records or directly measured using the Boehringer Manheim cholesterol enzymatic kit (Roche Diagnostics, Indianapolis, IN, USA) as previously detailed [10]. Given that LDL-C measurements derived from medical records were estimated using the Friedewald equation [12], any individual with triglyceride levels greater than 400 mg/dL were coded as missing for LDL-C. LDL-C measurements greater than four standard deviations from the mean were coded as missing in order to exclude undue influences of extreme outliers. CATHGEN non-familial cohort The CATHGEN biorepository consists of sequential individuals recruited through the cardiac catheterization laboratory at Duke University Hospital (Durham, NC). Case-control status in the CATHGEN sample was assigned as a function of coronary artery disease index (CADi) and age, such that controls had reached a sufficient age

7523047

to be at risk of developing disease [9]. All CATHGEN subjects fasted for a minimum of seven hours prior to blood sample collection. Blood was collected via the femoral artery and processed immediately for collection of plasma, and then frozen within hours at -80°C. For this study, we selected 1908 CATHGEN subjects based on their CAD status as previously used for genetic analyses [9]. We had two sources of LDL-C levels; derived from the medical records for a subset of the participants; and lipoprotein particle number concentration measured in stored, frozen, fasting plasma by nuclear magnetic resonance spectroscopy through Liposcience (Liposcience, Raleigh, NC, USA), using published techniques [13,14]. For quantitative trait analyses, total LDL particle number (LDLP) was used as a surrogate for LDL-C; in those CATHGEN individuals for whom both LDL-C and LDLP levels were available (N = 669), the two measures were strongly correlated (r = 0.67, p < 0.0001). Leptin levels were available for 380 individuals from the CATHGEN sample, as previously reported [10,15] Human Aorta Tissue Collection A collection of 162 aortas were harvested from deceased donors and prepared as previously described [16]. DNA was extracted from the tissue for genotyping using standard protocols. In addition, histopathological studies of the aortas were performed. Specifically, samples were assessed for extent of early atherosclerotic lesions with Sudan IV staining and severe disease assessed by the extent of raised lesions. The burden of atherosclerosis in the aortas was measured using the protocol described in the Pathobiological Determinants of Atherosclerosis in the Young study (PDAY) and

7523047

were given a graded score (1-4) [17]. As these aortas were harvested from deceased donors, the clinical information attached to each sample was limited, consisting of sex, age, and race. SNP Selection SNPs were selected for genotyping based on both the physical distance between SNPs (density dependent selection) and based on the pattern of linkage disequilibrium (LD) within the region (tagging SNPs). SNP map positions and gene identities were derived from the most recent draft of the human genome available (GRCh37/hg19). Within the region of highest lod scores, the average SNP spacing was 1 per 20 kb. Within the flanking genomic regions, SNPs were selected for an average density of 1 per 50 kb. In addition tagging SNPs were selected to capture LD information within coding regions using HapMap data and the Tagger algorithm with the following criteria: an r-squared greater than or equal to 0.7 and a minor allele frequency (MAF) greater than or equal to 0.05; LD between SNPs was visualized using Haploview [18]. One SNP was chosen for each LD bin. Priority was given to coding SNPs, followed by SNPs within known regulatory regions, intronic SNPs, and SNPs located within the 5' or 3' UTR, resulting in a list of 744 haplotype tagging SNPs with one SNP per LD bin. Finally, any SNPs with an Illumina score of less than 0.6, a MAF of less than 0.05, or coded as a potential failure by the Illumina software were excluded from the selection. The final list contained 2,256 SNPs, composed of both density-dependent (1,512)

7523047

and tagging (744) SNPs. Genotyping Genomic DNA from the GENECARD and CATHGEN samples was extracted from whole blood using the Pure-Gene system (Gentra Systems, Minneapolis, Minnesota, USA). SNPs were genotyped in two rounds, initially at the Center for Human Genetics at Duke University and subsequently through the NHLBI funded Seattle SNPs (http://pga.gs.washington.edu). The genotyping was performed using the Illumina GoldenGate technology (San Diego, CA, USA). To ensure genotyping accuracy and reliability, several quality control methods were used including two HapMap CEPH individuals and two duplicate individuals included per 96-well plate. SNPs with call rates less than 95% (N = 174) were excluded and individuals with a less than 90% genotyping rate were excluded (N = 57), resulting in 2082 SNPs on a total of 2823 individuals available for analysis. Of those, 20 deviated significantly from Hardy-Weinberg equilibrium (HWE) in Caucasians (p < 0.001) [19]. These 20 SNPs were analyzed, as it has been shown that some deviations from HWE are consistent with reasonable models for complex disease [20]. However, none of these 20 SNPs was significant in any of the analyses performed and had no impact on the reported results. Statistical Analyses Association with Quantitative LDL-cholesterol Traits In the GENECARD sample, association between each individual SNP with LDL-C was performed using the quantitative disequilibrium test (QTDT) [21] and a linear model. Of the available GENECARD sample, an average of 122 trios was analyzed per marker (range 30-196, median 130). In the CATHGEN sample, genotypic and allelic associations between each individual SNP and LDLP were assessed using

7523047

multivariable logistic regression adjusted for race, age, and sex. Given the low power for association in the GENECARD study, we chose to combine the quantitative analysis results for each SNP from GENECARD and CATHGEN using Fisher's method for combining p-values. Given previous reports that one of our identified genes (EBF1) is associated with decreased leptin levels in a murine knockout model [22], we used the Wilcoxon rank sum test to test for association of SNPs within and flanking EBF1 with leptin levels in those CATHGEN individuals with available leptin data (N = 380). Association/Linkage Analysis with Cardiovascular Disease The total SNP panel was tested for association with CAD in GENECARD and CATHGEN and with atherosclerosis in the aorta samples. In the GENECARD study, parametric two-point linkage for early-onset CAD was performed with a recessive (at risk allele freq 0.20, and penetrance 0.001) and dominant (at risk allele freq 0.01, and penetrance 0.001) model using Vitesse [23] and Homog [24]. These tests were conducted to provide independent validation of SNPs showing evidence for association. In the presence of association and linkage, there tends to be a positive correlation between tests of association and linkage; however, there is no such correlation between tests when linkage and association are not present, implying that true positive results in one test tend to be reflected by positive results in another [25]. Family based association with early onset CAD was performed using the association in the presence of linkage test (APL) [26]. This test appropriately accounts for the non-independence of affected siblings

7523047

and calculates a robust estimate of the genetic variance. In the CATHGEN cohort, we used multivariable logistic regression adjusted for race and sex, using allelic and genotypic models to test for association with CAD casecontrol status. In addition, a second CAD case-control series was constructed using the subset of the GENE-CARD probands (N = 150) that were sampled from North America to more closely resemble the CATH-GEN controls (N = 400), as we have previously done [9]; in these analyses, logistic regression was used to test for association between individual SNPs and CAD adjusted for race and sex. In the aorta sample, the qualitative phenotype was atherosclerosis status, defined by a histopathologic index of atherosclerosis [9], which was analyzed via multivariable linear regression, with adjustments for race and sex under genotypic and allelic models. For all analyses, any results with a nominal p-value ≤ 0.05 were considered to be significant. We did not specifically adjust for multiple comparisons, as our method of comparing multiple related analyses in independent datasets, which we refer to as analyses in parallel, provides internal replication for significant results in the same gene in multiple analyses and further support the significance of the initial observation. Analyses were conducted using SAS Version 9.1 (Cary, NC) unless otherwise specified for specific statistical analysis programs (i.e. QTDT, APL and linkage). Quantitative Trait Associations for LDL-cholesterol Traits The baseline characteristics of the overall study populations have been reported [9]. LDL-C measurements were available for 827 individuals in the GENECARD cohort, with a mean LDL-C concentration of

7523047

127.1 mg/dL (standard deviation [SD] 52.3 mg/dL). Family-based association resulted in 32 SNPs in 17 distinct genes that were significantly associated with LDL-C (Additional File 1 Table S1), with the four most significant SNPs residing in the gene PRELID2 , (PRELI domain containing 2 isoform) (rs10074645, rs6893183, rs17103583, and rs1865009, p = 0.0002-0.002), which were all in linkage disequilibrium with each other (D' ranging from 0.85-0.95). In the CATHGEN cohort (N = 1,908), mean LDLP levels were 1,131 nmol/L (SD 413 nmol/L). In CATHGEN, we found 102 SNPs in 46 distinct genes that were significantly associated with LDLP levels (Additional File 2 Table S1), with the most significant findings for SNPs in the genes SPOCK1 (sparc/osteonectin, cwcv and kazal-like domains) (rs17600115, p = 0.003) and PPP2R2B (phosphatase 2 regulatory subunit B family) (rs2125443, p = 0.005). Given low power for association in the GENE-CARD study, the p-values from the GENECARD and CATHGEN studies were combined using Fisher's method, resulting in 51 SNPs with combined p-values ≤ 0.05 (Table 1). As such, we found ten genes with significant results for association with LDL-C phenotypes in both the GENECARD and CATHGEN cohorts, with the strongest, most consistent results for SNPs in PRELID2, SPOCK1, PPP2R2B, and EBF1 (Early B-Cell Factor 1). The results for the quantitative analyses with LDL-C traits are summarized in Figure 1. Given previous studies showing that mouse knockout models for EBF1 have reduced leptin levels [22], we also tested all EBF1 SNPs and SNPs upstream and downstream of EBF1 (N = 78 SNPs) for association

7523047

with leptin levels (median 13.8 micrograms/L, range 0.4-104.9 micrograms/L) in a subset of CATHGEN (N = 380); six EBF1 SNPs were nominally significantly associated with leptin levels (Additional File 3 Table S1). The most significant result was for rs13165442 (p = 0.001), however that SNP was not significant for association with lipid levels or CAD. The SNP rs17635991 in EBF1, however, was nominally associated with CAD (p = 0.02) and with leptin levels (p = 0.03). Qualitative Trait Association with CAD/atherosclerosis Those genes with significant results for association with LDL-C traits were retained for comparison to the CAD endpoint results; only those genes which had a significant result in at least one CAD endpoint were retained for analysis. This list was further reduced by selecting those genes with the lowest p-values and the most consistent results across all analyses (i.e. significant results for the largest number of independent tests). Analyses in Parallel Of the 51 SNPs in ten genes that were significantly associated with LDL-C traits, we further reduced this list of candidates by selecting only those genes which had at least one significant association with any of the CAD endpoints, resulting in a list of nine genes ( Figure 3). These genes were then prioritized by both the size of the p-value and consistency in effect across both quantitative and qualitative analyses. This selection resulted in a final list of four genes (Table 2) with the most consistent association with LDL-C traits and CAD/atherosclerosis, although with varying effects for SNPs within those genes. The LD

7523047

pattern for significant SNPs within each gene showed few examples of strong LD between significant SNPs (Additional Files 9, 10, 11 and 12), with the exception of PRELID2 (Additional File 12). In order to verify the independence of these four loci, pairwise LD was calculated for all SNPs across the four genes; there was no LD between SNPs from differing genes (data not shown). Discussion We have demonstrated herein that disease-related intermediate traits can identify novel disease risk genes. Specifically, we used LDL-cholesterol traits to fine-map a linkage peak on chromosome 5 from the GENECARD study of early-onset CAD with integration of these results with association and linkage to cardiovascular disease. Using this approach, we have identified four candidate genes (EBF1, PPP2R2B, PRELID2, and SPOCK1) that may represent novel cardiovascular disease risk genes mediated through LDL cholesterol pathways. Although no genome scans for CAD or MI have reported linkage to this region, there are several potentially related phenotypes that have been mapped to the 5q31 locus including inflammatory or autoimmune conditions (celiac disease [27], asthma [28], Grave's disease [29], psoriasis [30], and Crohn disease [31]) as well as cardiac and vascular phenotypes (cardiomyopathy [32], intracranial aneurysm [33], infantile hemangioma [34], and systolic and diastolic blood pressures [35,36]). Several GWAS and meta-analyses have been published for lipid related traits [37][38][39][40]. However, only one study has reported a significant association for any genes on chromosome 5q31-33, for the gene T-cell immunoglobulin and mucin domain containing 4 (TIMD4) [38], which is 1.7 Mb centromeric to EBF1 and 9.9 Mb

7523047

telomeric to PPP2R2B. Nominally significant results were obtained for SNPs within TIMD4 and its ligand (HAVCR2), which is 152 kb telomeric to TIMD4 (Additional Files 2 Table S1, 6 Table S1, and 8 Table S1). Using a database of publicly available GWAS results from NHGRI, we looked for any reports of significant associations for SNPs within the 1 lod interval on chromosome 5. There are few references to cardiovascular disease or disease-related traits. However, none of these SNPs overlap with significant SNPs in our results, with the exception of two studies of hypertension/systolic blood pressure [41,42], each of which reported one significant association with a SNP in EBF1; unfortunately, neither of those two SNPs was examined in our study. The gene EBF1 is involved in hematopoiesis and immunity [43]. Interestingly, studies of knockout mice have identified a role for EBF1 in metabolism [22], a cardiovascular disease related phenotype. The null mice described by Fretz et al. have a unique metabolic syndrome characterized by lipodystrophy, hypotryglyceridemia, and hypoglycemia, while having an increased metabolic rate and decreased leptin levels. The mouse lipodystrophy is characterized by an increase in yellow adipose tissue in bone marrow and a marked decrease in white adipose tissue (by as much as 90%), relative to wild type controls. These findings are consistent with EBF1's regulation of adipocyte progenitors [44,45]. In our study, SNPs in EBF1 were significantly associated with LDL cholesterol traits and the CAD endpoints, with the exception of two-point linkage. In addition, EBF1 variation was associated with leptin levels in our sample,

7523047

although the results for individual SNPs were inconsistent with their association with lipids and CAD endpoints. This may suggest that EBF1 has a similar role in regulating adiposity and lipid metabolism in humans, and that variants in the gene may represent good candidate polymorphisms for cardiovascular disease and dyslipidemia in humans. Of the other candidate genes identified, SPOCK1 is associated with age at menarche via a genome-wide association study [46,47]. SPOCK1 encodes a proteoglycan that functions as a protease inhibitor; although initially identified in testes [48], it is expressed in many human tissues including blood. SNPs within SPOCK1 were tested for sex-specific effects in our sample via a stratified analysis; no significant sex effects were observed (data not shown). PPP2R2B encodes a brain specific regulatory subunit of a protein phosphatase and is the causal locus for a Mendelian disease, a form of spinocerebellar ataxia (SCA12, OMIM# 604326). Little is known about the function of the gene PRELID2 other than it contains a 'prel-like domain,' from which its name is derived. Using a strategy of analyses in parallel we have identified four novel candidate genes for cardiovascular disease. The ability to reduce the list of potential candidates within the linkage region on chromosome 5q31-33 from a few hundred to only four is proof of principle that this strategy may be a useful tool for analyzing complex traits. In addition, had we relied upon CAD endpoint analyses alone, we would have obtained less significant associations overall and would have prioritized a different set of candidate genes. One of

7523047

the major strengths of our study is the detailed phenotype information available for both the GENECARD cohort and CATHGEN biorepository. The rigorous inclusion criteria and case definitions used in GENECARD and CATHGEN have led to objective measures for CAD endpoints, a phenotype that would otherwise have a subjective definition. In this particular study, direct sequencing, rather than SNP genotyping across multiple samples, would not have been appropriate; if we were to re-sequence a sub-set of the sample, it is not clear which individuals would be selected for such sequencing, particularly for the continuous quantitative traits we examine. Finally, our sample is of mixed ethnicity (Caucasian and African-American), which would necessitate a separate re-sequencing effort for each ethnicity. This study has some limitations. First, all four genes may be CAD susceptibility genes and act independently (as the LD patterns in our samples suggest). However, it is possible that SNPs within our sample are in LD with causal SNPs at another locus and the association results from these four genes may not be independent. Second, we have focused on the consistency of the results at the gene level (i.e., which genes have SNPs that are significant in multiple analyses). However, it is not the case that the same SNPs are significant in those analyses or, in the case that the same SNP is significant, that the magnitude of that significance is similar between analyses. Thus, we cannot begin to identify individual SNPs within a candidate gene that are likely to be driving the results via direct or indirect

7523047

biological action. This can be explained, in part, by the fact that the phenotypes, while correlated, are not perfectly correlated. Therefore, it is expected that there will be differences in the pvalues for associations with different phenotypes and this could cause certain associations to fall outside the nominal p-value cutoff. Finally, the results were not interpreted in the context of correction for multiple Figure 3 Analytical Strategy for 'Parallel Analysis' of Chromosome 5q31 Region. This figure details the study design, cardiovascular cohorts and analytic techniques used, and the number of unique genes containing SNPs with significant associations (indicated underneath the method used). The union of quantitative results (α) indicates the subset of genes shared by the two methods. The qualitative trait (CVD) was analyzed in GENECARD, CATHGEN, and aortas and the total number of unique genes containing a significant SNP in any of those analyses is indicated. The commonality of the genes between the quantitative and qualitative analyses (N = 9, PRELID2, SPOCK1, EBF1, PPP2R2B, DMXL1, DTWD2, GABRG2, GLRA1, and RP11-166A12.1) is indicated by ω. comparisons. There are two main difficulties with applying such corrections to these results, i.e. a Bonferroni correction which would be overly conservative. First, the phenotypes examined are correlated, therefore the analyses conducted using more than one phenotype within the same sample are not independent. Second, if we look at the results sequentially, with one analysis conducted after another, then the prior probability that a given SNP in a gene of interest will be significant in subsequent analyses is non-negligible. We are

7523047

not relying upon the magnitude of any given p-value to identify a single gene in the region as the most likely to explain the original evidence for linkage. Rather, we are suggesting that a set of genes be examined as likely candidate susceptibility loci for cardiovascular disease that is mediated by lipid levels. In order to identify which gene or genes among the four we have selected contains variation for cardiovascular disease mediated by LDL cholesterol pathways, there are several methodologies available. Re-sequencing studies could be conducted in our sample, either in the entire population or by using individuals with extreme trait values (i.e., very high/low LDL-C levels or very early onset cardiovascular disease), as these data would capture all of the variation present in those samples and not rely upon common variants identified through a different, although ethnically similar, sample (i.e., CEPH Caucasians). In addition, those genes for which the biological function is known (i.e., EBF1) could have their level of activity or functionality directly assessed in genotyped samples. Such an approach can identify subsets of variation that appear to have functional consequences. However, due to LD within the variants, such results can still be ambiguous, in which case promoter and gene constructs can be created and assayed in the laboratory, allowing one to query the functional consequences of individual variants Conclusions In summary, we propose a strategy of parallel analysis where results from analyses of disease-related intermediate traits for a complex disease can be considered jointly with qualitative results from mapping of the disease

7523047

trait itself, potentially enabling discovery of novel disease genes mediated through these intermediate phenotypes that might not have been identified using disease status alone. In addition, this strategy may allow the dissection of genetic heterogeneity mediated through the intermediate phenotype. We have applied this strategy to the fine mapping of a linkage peak for early-onset CAD and thereby have demonstrated replication of four genes within a region on chromosome 5q31. These genes, in particular, EBF1 given its potential biological plausibility, serve as novel candidate loci for cardiovascular disease and should be further evaluated.

7523047

Changes in distribution of nuclear matrix antigens during the mitotic cell cycle. We examined the distribution of nonlamin nuclear matrix antigens during the mitotic cell cycle in mouse 3T3 fibroblasts. Four monoclonal antibodies produced against isolated nuclear matrices were used to characterize antigens by the immunoblotting of isolated nuclear matrix preparations, and were used to localize the antigens by indirect immunofluorescence. For comparison, lamins and histones were localized using human autoimmune antibodies. At interphase, the monoclonal antibodies recognized non-nucleolar and nonheterochromatin nuclear components. Antibody P1 stained the nuclear periphery homogeneously, with some small invaginations toward the interior of the nucleus. Antibody I1 detected an antigen distributed as fine granules throughout the nuclear interior. Monoclonals PI1 and PI2 stained both the nuclear periphery and interior, with some characteristic differences. During mitosis, P1 and I1 were chromosome-associated, whereas PI1 and PI2 dispersed in the cytoplasm. Antibody P1 heavily stained the periphery of the chromosome mass, and we suggest that the antigen may play a role in maintaining interphase and mitotic chromosome order. With antibody I1, bright granules were distributed along the chromosomes and there was also some diffuse internal staining. The antigen to I1 may be involved in chromatin/chromosome higher-order organization throughout the cell cycle. Antibodies PI1 and PI2 were redistributed independently during prophase, and dispersed into the cytoplasm during prometaphase. Antibody PI2 also detected antigen associated with the spindle poles. The nuclear matrix is a complex biochemical fraction consisting of nonhistone nuclear proteins and small quantities of DNA and RNA . It is obtained by sequential extraction

386312

of isolated interphase nuclei with low-and high-salt buffers, detergents, and DNAse and RNAse . Structurally, the nuclear matrix comprises the peripheral nuclear pore complex-lamina, an internal fibrogranular network and residual nucleoli. Part of the matrix has been envisaged as an interphase "nuclear skeleton" on which nuclear functions such as DNA replication, RNA transcription and processing, virus replication, and hormone response can be ordered (see references 1-3 for reviews). Several studies have suggested that the nuclear matrix is involved in mitosis (4)(5)(6)(7). Recent work (8,9) has also indicated a potential role for the nuclear matrix in the organization of mitotic chromosomes per se. To examine the distribution of individual nuclear matrix polypeptides throughout the cell cycle, we used monoclonal antibodies against isolated applied for 30 min; the c .s. were rinsed in phosphate-buffered saline (PBS), pH 7 .0, flooded with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse Ig for 30 min, washed in PBS, and mounted in 50% glycerol-PBS with 0.1% p-phenylenediamine to retard fluorescence bleaching. Immunoblotting: Lymphocyte nuclear matrix proteins were separated in SDS polyacrylamide gels according to Brasch (12) and were transferred electrophoretically to nitrocellulose (Bio-Rad Laboratories, Richmond, CA) by a modification of the method of Burnette (13). The electrode buffer contained 192 mM glycine and 25 mM Trizma base. The total transferred proteins were visualized directly by staining a strip of the nitrocellulose sheet with amido black. Molecular weights were estimated by comparison with high-and lowmolecular-weight standard kits (Bio-Rad Laboratories, Mississauga, Ont .) . Immunofluorescent Staining : Mouse 3T3 fibroblasts were grown on c .s. and

386312

fixed in 3% paraformaldehyde-PBS, pH 7 .0, for 5 min at room temperature . They were reduced in sodium borohydride in PBS, permeabilized for 20 min in 0 .2% Triton X-100 in PBS, washed in PBS, drained, and flooded with 1°antibody . Control samples were flooded with PBS . The c .s. were incubated for 45 min, drained, washed in PBS, flooded with 2°antibody, and incubated for 45 min . After washing in PBS, the c.s . were placed in Hoechst 33258 (American Hoechst Corp ., San Diego, CA) (1 .5 Vg/ml in PBS, pH 7.0) and then mounted in 50% glycerol-PBS with p-phenylenediamine. Double Immunofluorescence Staining: After fixation, permeabilization and washing in PBS as above, the c .s, were placed for 20 min in 0.15% gelatin in PBS and washed three times for 10 min in PBS. The antibodies were then applied as follows : first 1°antibody, 45 min ; three times for 4 min in PBS; first 2°antibody, 45 min ; three times for 4 min in PBS ; second 1°a ntibody, 45 min ; three times for 4 min in PBS ; second 2°antibody, 45 min ; three times for 4 min in PBS ; 2 min in Hoechst 33258 ; mount in glycerol-PBS . Antihistone (1 :200) was a previously characterized serum from a patient with an autoimmune disease, kindly supplied by Dr. M. J . Fritzler, University of Calgary. Antilamin (1 :400) was a serum from a patient with scleroderma, provided by Dr. M . Kirschner (14). Microscopy : Preparations were

386312

observed with a Zeiss Universal microscope (Carl Zeiss, Inc ., New York) equipped for epifluorescent illumination and were photographed on Ilford XP 1-400 film . Selection of Monoclonal Antibodies We selected 17 hybrids producing antibodies that detected nuclear antigens . The staining patterns of interphase nuclei of 3T3 cells with these antibodies fell into three categories : peripheral (P), internal (I), or both (PI) . In this study, we report observations using antibodies P1 and 11, produced against bovine retropharyngeal lymphocyte nuclear matrix, and antibodies PI 1 and PI2, produced against mouse splenic lymphocyte nuclear matrix . In an earlier report (15), these antibodies were referred to as 3F5, 5H12, 1G11, and 2F8, respectively. All four antibodies were detected by FITC-conjugated goat anti-mouse IgM, but not by FITC-conjugated goat anti-mouse IgG. The structures and the polypeptide compositions of the isolated lymphocyte nuclear matrices have been described 66 2 THE JOURNAL OF CELL BIOLOGY -VOLUME 99, 1984 previously (10,16) . Fig. 1 shows immunoblotting results with lymphocyte matrices for three of the four monoclonal antibodies used in this study . Antibody P1 detected a triplet of polypeptides in the range of M, 27,000-31,000 and PI 1 a single polypeptide of Mr 60,000 . Antibody PI2 detected three components (M, 35,000, 70,000, and 140,000), which may be related subunits or degradation products . We have been unsuccessful so far in identifying the antigen to I1 by immunoblotting. Interphase The antigen detected by PI was exclusively peripheral in interphase (Fig. 2, a-c). Focusing on the midplane of the nucleus

386312

showed a relatively uniform staining of the periphery (Fig . 26), with some small irregular projections towards the interior of the nucleus. Focusing on the surface of the same nucleus (Fig . 2 a) showed that the projections were uneven in size and apparently distributed at random . The staining pattern with antibody PI was compared to that with antilamin ( Fig . 3, a and b). Peripheral staining was very prominent with antilamin, but showed no internal projections and was agranular. Furthermore, there was a significant fluorescence in the nuclear interior. The internal staining with antilamin consisted of diffuse non-nucleolar fluorescence and some bright irregular streaks and/or spots . Antibody I1 stained exclusively the internal region of the nucleus . The stain appeared as specks of fairly uniform size Interphase nuclei stained with P1 (a and b), 11 (d), PI1 (f and g) or PI2 (i and j) antibodies, and counterstained with Hoechst (c, e, h, and k, respectively) . Antibody 11 stained only the nuclear interior (d) (the arrowhead points to a nucleolusassociated brightly stained mass) . Selected focal planes showing nuclear surface (a, f, and i) and interior (b, g, and k) are depicted for the other antibodies . Hoechst micrographs (c, e, h, and k) are at the same focal plane as the micrographs of the interior (b, d, g, and j, respectively) . x 1,450 . FIGURE 3 Cells in interphase and prometaphase stained with antilamin antibody (a) and Hoechst (b) . The interphase nucleus shows intense peripheral staining, some brightly

386312

stained internal spots and a diffuse internal stain ; nucleoli (arrowheads) are not stained . In prometaphase, fluorescent material is visible in some areas at the periphery of the chromosome mass and the cytoplasm is brightly stained . x 1,450 . and intensity distributed in the regions between nucleoli and heterochromatin bodies, i .e ., the interchromatinic region ( Fig. 2, d and e). Occasionally, larger bright spots were seen associated with the edge of one or more of the nucleoli. With antibody PI 1, the staining pattern in most nuclei was both peripheral and internal (Figs . 2, f-h, and 4, a-d) . In some cases, however, only peripheral staining was observed (Fig. 4, c and d) . The peripheral staining was similar in both situations . Viewed in the midplane of the nucleus (Fig. 2g), the peripheral staining consisted of a discontinuous layer of uniform thickness . In surface view (Fig. 2f), this staining was resolved into evenly distributed coarse granules of uniform size and brightness . The internal staining component was unevenly distributed throughout the nucleoplasm as numerous granules that varied greatly in size and brightness . There was no obvious staining of nucleoli or heterochromatin . A further characteristic of PI 1 staining was the presence in the cytoplasm of some cells (Fig. 4, a and b) of a cluster of brightly staining granules of variable size near the nuclear surface. The interphase staining pattern with antibody P12 was similar to that with PI 1 (Fig. 2, i-k) . However, there was

386312

little variation in the staining pattern from one cell to another, and the staining was exclusively nuclear in all interphase cells. The peripheral staining consisted of finer, more irregularly sized granules than with PI 1, as is clearly seen in Fig. 2 i. The internal staining was also somewhat finer ( Fig. 2j) and appeared to be more evenly distributed than with PI 1 . Mitosis Cells were identified as to mitotic stage, using primarily DNA staining with Hoechst, according to the following criteria: (a) prophase (Figs. 5 and 6)-from the first sign of chromosome condensation up to the loss of a smooth nuclear outline resulting from the start of nuclear envelope breakdown ; (b) prometaphase-from nuclear envelope breakdown to the formation of the metaphase plate ; at mid-prometaphase ( Fig. 7), chromosomes were oriented radially around a . The same nuclei counterstained with Hoechst are shown in c, f, and i, respectively . Surface invaginations are more numerous in the P1-stained nucleus than at interphase (cf .a, 1 a) . Two groups of cytoplasmic granules stained by PI1 can be seen near the nuclear surface (d) . The peripheral staining by PI2 (h) is diffuse, and the surface shows some "wrinkling" (arrowheads) . The internal staining by PI 1 (d and e) and PI2 (g and h) is patchy . x 1,450. 9)-the daughter cells were still connected through the midbody when viewed by phase-contrast microscopy . The paired nuclei were smaller than in surrounding interphase cells, but were otherwise similar in appearance, with a

386312

diffuse DNA fluorescence throughout the nucleoplasm and well-defined nucleoli . It should be noted that no squashing was applied to the cell preparations so that the arrangement of chromosomes remained largely undisturbed from that present in living cells. Prometaphase nuclei stained with P1 (a), 11 (c), PI1 (e), or PI2 (g) antibodies and counterstained with Hoechst (b, d, f, and h, respectively) . (a and b) The chromosomes (b) are oriented radially around a central space with the centromeres (arrows) (see also g-i) towards the center . Antibody P1 stains the periphery of the chromosome mass, coating the projecting chromosome arms . (c and d) Antibody 11 is distributed as specks along the surface of the chromosome (arrowheads) . The interior of the chromosome also shows some diffuse fluorescence . (e and f) Antibody I'll is not chromosome associated . Faint granular fluorescence is visible in the central space of the prometaphase configuration as well as between some chromosomes . (g and h) Antibody PI2 is not chromosome associated . Two foci of stain are evident, resembling the staining of involutions of the nuclear periphery observed in prophase (cf. Fig. 4g) . x 1,450 . Prophase In prophase, the staining intensity with antibody P1 increased manyfold, so that adjacent interphase and prophase nuclei could not be photographed with the same exposure . This elevated staining intensity occurred throughout mitosis . The stained irregular inward projections from the nuclear periphery present in interphase were more numerous in early prophase (Fig. 5, a-c) . This was seen most

386312

clearly when the inner surface of the nuclei was examined (Fig. 5 a). In later prophase (Fig. 6, a and b), the projections were seen to coincide with the surface of short lengths of condensed chromosomes, coating the chromosomes where they lay at the nuclear periphery . Although condensed chromosomes were evident throughout the entire nucleus by DNA staining (Fig. 6 b), only those portions at the periphery of the nucleus were coated with antigen to Pl . All chromosomes at the nuclear periphery showed surface staining with P 1 . The internal staining of interphase nuclei with antibody I1 was redistributed at prophase (Fig. 6, c and d), apparently associating itself with the condensing chromosomes. In contrast to the staining pattern of P1 (Fig. 6, a and chromosome segments throughout the nucleus were stained with 11 . The stain was distributed as specks along the entire surface ofeach chromosome segment, and there also appeared to be some diffuse staining of the interior of each chromosome. Early in prophase the staining of the nucleoplasm with antibody I'll (Fig. 5, d-f) was patchy, indicating clustering of the interphase specks of stain between condensing chromosomes. There was no evidence of any direct association of stain with the chromosomes. Brightly staining granules were observed in the cytoplasm of some prophase cells, usually in two clusters near the surface of the nucleus (Fig. 5, d and e). The peripheral nuclear staining was accentuated later in prophase (Fig. 6, e and f), where two involutions of the nuclear periphery were readily

386312

visible (Fig. 6, e andf). The internal component staining with P12 became redistributed early in prophase (Fig. 5, g-i), lying in patches between condensing chromosomes. The peripheral staining was no longer granular (Fig. 5 h), and the surface appeared "wrinkled" in places . Later in prophase (Fig. 6, gand h), internal staining disappeared and the peripheral staining consisted of bright diffuse fluorescence . As observed with PI1 (Fig. 6, e and f), the nuclear periphery was involuted in two places at prophase (Fig. 6, g and h). No staining of cytoplasmic granules was seen with PI2 . Prometaphase At prometaphase, P1 was detected only at the periphery of the chromosome mass (Figs . 7 a and 10, a, c, and e), closely following the outlines of projecting chromosome arms (Figs. 7 b and 10, b, d, and f). No stain was detected within the individual chromosomes or within the chromosome mass. This is illustrated in the through-focus series in Fig. 10, a-f. All chromosome arms at the surface ofthe configuration (Fig. 10, b andf) were coated (Fig. 10, a and e), but in the midplane of the mass, only those portions of the chromosomes at the periphery (Fig. 10 d) were coated by P1 (Fig. 10 c). In contrast, prometaphase chromosomes in preparations stained with antilamin were unstained (Fig. 3, a and b). The cytoplasm showed considerable diffuse fluorescence, and in some cells granular bright lines could be seen in places in the cytoplasm around the chromosome mass. To ascertain that peripheral staining of the chromosome

386312

mass by P1 was not simply due to exclusion of antibody, we performed double immunostaining with P1 and antihistone ( Fig. 10, g-h). Although the chromosome mass was only outlined by P1 (Fig. 10g), each chromosome was fully stained by antihistone (Fig. 10h), resembling the pattern recently described with a monoclonal antibody to histone H2b (17). With antibody 11, a finely granular stain was associated with the surface of the chromosomes at prometaphase (Fig. 7, c and d). There also appeared to be some diffuse staining of the interior of the chromosomes themselves, as had been seen in prophase nuclei (Fig. 6, c and d). Antibodies PI 1 and PI2 were not chromosome associated after nuclear envelope breakdown . Staining with both antibodies was greatly reduced at prometaphase (Fig. 7, e-h). Some specks of stain were visible between chromosomes and in the central space of the prometaphase configuration, perhaps corresponding to the sites of involution of the nuclear periphery observed in prophase . Metaphase, Anaphase, and Telophase At metaphase (Fig. 8, a and b), the periphery ofthe metaphase plate was precisely outlined by P 1 (Fig. 8 a), with no visible staining of the central portions of the chromosomes. However, any chromosomeslying outside the metaphase plate were individually delineated by P1 (Fig. 8a). In early anaphase, ifthe two daughter sets ofchromosomes were intertwined, P1 outlined them as one mass. At a more advanced stage of anaphase (Fig. 8, iJ), when the distal ends ofthe chromosome sets have moved past one another towards the poles, each daughter

386312

set of chromosomes was outlined separately by P1 (Fig. 8 i). During telophase, the distribution of P1 paralleled the surface irregularities of the condensing chromosome mass (Fig. 8, q and r). In many cells in telophase (Fig. 8 q) and in early G 1, there were brightly staining globular masses in the cytoplasm . With antibody 11, the metaphase (Fig. 8, c and d), anaphase (Fig. 8, k and 1) and telophase (Fig. 8, s and t) chromosomes were associated along their entire length with bright specks of stain as described for prometaphase. The interior of the chromosomes also showed a diffuse fluorescence . The PI 1 staining was reduced to a diffuse glow of the cytoplasm at metaphase (Fig. 8, e and f) and anaphase (Fig. 8, m and n). As the chromosomes began to lose their individual definition in telophase, a very faint stain could be seen at the periphery of the chromosome mass (Fig. 8, u and v), initially at the side of the mass nearer the spindle poles. FIGURE 8 Cells in metaphase (a-h), anaphase (i-p), and telophase (q-x) stained with P1 (a, i, and q), 11 (c, k, and s), PI1 (e, m, and u) or PI2 (g, o, and w) antibodies, and counterstained with Hoechst (b, i, and r; d, I, and t; f, n, and b ; h, p, and x, respectively) . Metaphase : Antibody PI stains the periphery of the metaphase plate (a and b) while 11 is seen as fine specks along the chromosomes (c and

386312

d) . With PI1 (e and f) and PI2 (g and h), the cytoplasm exhibits fluorescence, but no stain is associated with the chromosomes themselves . In cells stained with PI2, stain is also visible in two spots corresponding to the location of the spindle poles (g) . Anaphase : Antibody P1 outlined each daughter set of chromosomes (i and i) . Antibody 11 is associated with the chromosomes of each daughter set along their entire length (i) . The granular nature of the stain, its localization at the surface of the chromosome and the presence of some internal stain are particularly clear in this micrograph . The chromosomes are not stained by P11 (m and n) or PI2 (o and p) antibodies but the cytoplasm is fluorescent . Some PI2 staining is associated with the spindle poles . Telophase : Antibody P1 (q and r) is localized at the periphery of the chromosome groups, closely outlining the surface irregularities . Brightly staining globular masses are also visible in the cytoplasm (arrowheads) . Association of 11 with individual chromosomes is difficult to distinguish (s and t) ; the stain appears to be evenly distributed throughout the chromosome masses, and the speckling is less clearly evident . Antibody PI 1 (u and v) stains the surface of the chromosome masses on the sides closer to the spindle poles (arrowheads), as well as producing a diffuse cytoplasmic fluorescence . With PI2 (w and x), the cytoplasm is faintly stained as are the spindle poles . x 1,450 .

386312

Cells in early G1 stained with P1 (a), 11 (d), PI1 (g), or PI2 (j) antibodies, counterstained with Hoechst (b, e, h, and k, respectively), and photographed under phase-contrast illumination (c, f, i, and I, respectively) . The midbody is clearly visible in all cells . Although the nuclei are still somewhat irregular in shape (b, d, h, and k), the antibody-and Hoechst-staining patterns are similar to those in nuclei after cytokinesis (cf . Fig. 1) . x 1,450 . The chromosomes remained unstained with P12 during metaphase (Fig. 8, g and h), anaphase (Fig . 8, o and p), and telophase (Fig . 8, w and x) within a diffusely fluorescent cytoplasm . However, two stained spots located at sites appar-66 8 THE JOURNAL OF CELL BIOLOGY " VOLUME 99, 1984 ently corresponding to the spindle poles were visible during these stages (Fig. 8, g, o, and w) . At telophase, very faint staining was acquired at the periphery of the chromosome mass (Fig . 8 w) . In early G1, the daughter cells were still joined at the midbody (Fig . 9, c,f i, and 1) . Some cells showed the antibodyspecific staining pattern characteristic of interphase nuclei . Most, however, presented staining patterns intermediate between telophase and interphase . The staining with PI was somewhat less bright than earlier in mitosis but was still brighter than after cytokinesis. It was once more peripheral and fairly even in distribution (Fig. 9, a-c) . In parallel with the reappearance of interphase chromatin organization in early

386312

G1, the staining by antibody II was gradually redistributed throughout the nucleus as bright specks of even size (Fig . 9, df ). The peripheral components detected by monoclonals PI1 (Fig. 9, g-i) and P12 (Fig . 9, j-1) were clearly visible in all early G 1 cells . In nuclei showing a relatively diffuse Hoechst stain, similar to that in interphase nuclei (cf. Figs. 9, i and 1 and 1 c), the internal component was also evident (Fig . 9, g-1) . DISCUSSION The monoclonal antibodies studied here were generated against a biochemically characterized nuclear fraction, the nuclear matrix (10,16,18) . They were chosen because they localized antigens of interphase nuclei outside visible chromatin and nucleoli, in distributions paralleling the known location, from electron microscope studies, of different regions of the nuclear matrix, i .e ., the pore complex-lamina (PI), the fibrogranular internal matrix (I1), or both (PI1, PI2). Furthermore, we have demonstrated by immunoblotting that three of the four monoclonal antibodies detected polypeptides of the isolated nuclear matrix . Reports dealing specifically with nuclear matrix antigens are relatively few (19)(20)(21), although the distribution of the lamins during the cell cycle and cell differentiation has been extensively investigated (e .g. 14, [22][23][24][25][26][27] . Using the first monoclonal antibodies directed against nonlamin nuclear matrix antigens, we have demonstrated that, whereas the peripheral and internal regions of the nuclear matrix have some common antigens (detected by PI 1 and PI2), others (detected by P1 and 11) are restricted to one region or the other . We have

386312

also shown that, with the disruption or reorganization of the interphase nuclear matrix during mitosis, the antigens behave independently of one another . Some antigens are dispersed into the cytoplasm, as has been described for the lamins and most other nuclear antigens . Others, however, associate with the chromosomes in characteristic readily distinguishable patterns . These differences in behavior at mitosis may be a reflection of the particular nuclear matrix function in which the antigens are engaged during interphase . The observations with P1 and I1 support a role for the nuclear matrix in the spatial ordering of nuclear processes, a role maintained during mitosis as participation in the structuring of chromosomes . Monoclonals P1 and 11 were produced against bovine lymphocyte nuclear matrices, whereas monoclonals PI 1 and P12 were made against mouse lymphocyte nuclear matrices . All four antibodies detected antigens in the mouse 3T3 fibroblast cell line, as well as antigens in mouse and bovine lymphocytes . Indeed, antibody P1, for example, detects antigens with similar interphase and mitotic distributions in a number of mammalian, insect, and plant tissues (manuscript in preparation) . This suggests that the antigens under investigation are involved in basic nuclear functions common to many organisms . The antigen detected by P1 is localized at the nuclear periphery during interphase, and numerous studies have shown by immunocytochemistry that the lamins are associated with the nuclear periphery (14,22,23,26) . However, from our observations of both interphase and mitotic cells, it is quite clear that the antigens detected by the two

386312

antibodies are not the same . The interphase staining with P 1 is exclusively peripheral, while the antilamin reveals an internal component not only in our preparations but also in previously published reports (14,22) . At mitosis, whereas the P 1 antigen becomes associated with chromosomes, the lamins disperse into the cytoplasm . Furthermore, we have shown by immu-noblotting that P 1 detects an antigen of Mr 27,000-30,000 ; the lamins have been identified at Mr 60,000-70,000 (14,22) . Perichromin, a conserved nuclear envelope-associated antigen localized at the periphery of mitotic chromosomes has been recently detected by an autoimmune antibody (28) . The antigen has been identified as a single polypeptide of Mr 33,000 . Although períchromin is strongly peripheral at interphase in Chinese hamster ovary cells, it is uniformly distributed throughout the nuclei of Drosophila cells . The P 1 antigen appears to be distinct from perichromin . It is exclusively peripheral in all cell types examined, including insect cells (manuscript in preparation), and detects polypeptides of Mr 27,000-31,000. The monoclonal antibodies PI I and PI2 also detect components of the nuclear periphery in 3T3 cells . Taken together these data indicate that the nuclear periphery is more complex than previously considered . We propose that the antigen to P1 may function in the spatial ordering of DNA during interphase, and the spatial ordering of chromosomes during mitosis . Recently, a similar function has been proposed for perichromin (28) . Numerous workers have proposed that DNA is spatially ordered during interphase (29)(30)(31)(32)(33)(34), and that the

386312

order is established by anchoring of interphase DNA loops on the nuclear matrix (33)(34)(35)(36)(37) . At least some of these loops are anchored at the nuclear periphery (for reviews, see references 32 and 33) . The antigen detected by P l may be involved in anchoring DNA loops, by itself or by interaction with other nuclear proteins, e .g ., the lamins . There is also considerable evidence for spatial order of chromosomes during mitosis, i .e ., at the metaphase plate (32,38) . The behavior of the P1 antigen during mitosis is consistent with the notion that it may be implicated in translating interphase order into metaphase order . With the fragmentation of the nuclear envelope at the end of prophase, the P1 antigen, either by itself or in association with yet unknown factors, may remain chromosome associated so as to retain the spatial order present during interphase . It is unlikely that the P1 antigen interacts in a sequence-specific manner with the chromosomes . Considering only the regions of the centromeres, for example, it is clear that, whereas these are not coated by P1 at metaphase, they become associated with P1 during anaphase and telophase . The antigen to P1 may act merely as a physical constraint on the chromosomes, maintaining them in the relative positions established previously . It is interesting to note that Paulson (39) has shown that all the chromosomes from a cell adhere to one another and are isolated as one cluster under appropriate isolation conditions . He suggests that

386312

this may reflect suprachromosomal organization in vivo . The association of the II antigen with chromosomes during mitosis and its nonheterochromatin localization during interphase suggest that the antigen may be involved in structural organization of chromosomes both during interphase and mitosis . Since the I1 antigen appears to be localized largely at chromosome surfaces during mitosis, it is unlikely to be a constituent of the internal chromosome scaffold as prepared by Earnshaw and Laemmli (40). However, antigen 11 could be a component of the 25-ram granules recently demonstrated by Engelhardt et al . (41) in interphase nuclear matrices and chromosome scaffolds . The antigens for antibodies PI1 and PI2 showed similar distributions throughout the cell cycle, but with some characteristic differences . At interphase, the intercellular staining 670 THE JOURNAL OF CELL BIOLOGY -VOLUME 99, 1984 pattern was relatively consistent wtih PI2, but with PI 1, varied considerably in different nuclei . The type of staining displayed by a particular nucleus with PI I could not be correlated with any unique structure of the nucleus . It may be related to the metabolic state of the cell, perhaps to its position within the cell cycle . Riley and Keller (42) have shown extensive reorganization of a nuclear matrixlike fraction as HeLa cells progress through interphase . The interphase staining pattern as well as the dissolution and reacquisition of staining at prometaphase and telophase, respectively, suggest some association of the PI I and PI2 antigens with the nuclear envelope (33) . By mid-prophase, monoclonal PI2 showed only

386312

an agranular peripheral component. The altered appearance of the peripheral staining could be due to posttranslational modifications of the PI2 antigen similar to changes in the extent of phosphorylation of lamins and nuclear proteins which are known to occur at about this stage of mitosis (27,43,44) . It may also, however, reflect changes in distribution resulting from the loss (e.g ., the lamins) and/or gain (e .g., P1 antigen) of other peripheral proteins during prophase .

386312

Influencing factors of the Tmax parameter in Rock-Eval pyrolysis The rock pyrolysis parameter Tmax is an important index to evaluate the maturity of organic matter in source rocks. However, during the use of Tmax, it is found that the Tmax value of some samples does not conform to the thermal evolution law of organic matter when evaluating the maturity of organic matter, which affects the accuracy of geological applications. Studying the influencing factors of the rock pyrolysis parameter Tmax analysis data is beneficial to improve the experimental method to ensure the accuracy of the experimental data, and it is of great significance to the correct evaluation of the source rock type and maturity of the source rock. In this paper, the effects of sample mixing uniformity, sample weight, soluble organic matter and minerals on the rock pyrolysis parameter Tmax are analyzed experimentally. It is concluded that the degree of sample uniformity and the amount of sample will affect the results. Soluble organic matter will lead to a decrease in Tmax. With the change of kerogen/clay mineral ratio, the influence of clay mineral on the pyrolysis parameter Tmax. Introduction Organic carbon and rock pyrolysis analysis is one of the most basic analytical techniques in organic geochemistry research. For the reasons of rock pyrolysis parameters, the former has been studied. Zhang Zhenduo and others believe that the soluble organic matter enters the S2 peak, which leads to an increase in the S2 value and a decrease in the Tmax value. Clay minerals have an adsorption effect on heavy

210748914

hydrocarbons, resulting in a decrease in S2 value and an increase in Tmax. Liang Jimin and others believe that too small a sample size will result in a low S2 value and an increase in the Tmax value. For samples with low abundance of limestone and organic matter, a more accurate Tmax value can be obtained by changing the heating rate. By studying the influencing factors of the pyrolysis parameter Tmax, the influence of high soluble organic matter in source-storage source rocks on pyrolysis parameters can be clarified, and it is proved that the amount of pyrolysis hydrocarbons is closely related to the type and content of mineral matrix. The influence of Tmax on the pyrolysis of kerogen was determined to determine the range of sample sizes of three types of source rocks: mudstone, carbonaceous mudstone and coal in rock pyrolysis analysis. In this paper, 25 samples of low-to-high organic carbon content in the Junggar Basin were collected, and five basic analyses of organic carbon, pyrolysis, chloroform asphalt A extraction, group components, elements and isotopes were carried out on the rock samples. Appropriate rock samples 2 and kerogen samples were tested for rock pyrolysis, sample uniformity, sample weight, soluble organic matter, and clay minerals. Effect of sample uniformity on pyrolysis parameters Tmax Under normal circumstances, after the rock sample is selected by 100 mesh after being pulverized, it is analyzed by pyrolysis instrument. Whether the sample is evenly mixed and affects the pyrolysis results needs further analysis. In this paper, the samples from XJC1 well, DX8

210748914

well and C58 well were selected, and their TOC values were 1.55, 6.18 and 13.80. The three mudstone samples with low, medium and high organic carbon values were elected to rock pyrolysis experiments to analyze the sample uniformity and pyrolysis parameters. The samples were separately mixed for over 3 minutes using an XH-C type vortex mixer, and then each of them was couducted by parallel analysis of rock pyrolysis five times. The Tmax values of the mud samples in the XJC1 well before and after being mixed were all distributed between 470 °C and 472 °C, with an average of 471 °C. 471 °C was used as the standard value of the mudstone sample, and the Tmax value of the sample before being mixed needed deviation calculation. The maximum deviation of the Tmax value of the XJC 1 well was 1 °C. The mud samples' Tmax values of the DX8 well before being mixed were between 451 °C and 453 °C, and the average value was 452 °C. The Tmax values of the samples after being mixed were between 451 °C and 452 °C, and their average value was 451 °C. Take 451 °C as the standard value of the mudstone sample, and the Tmax value of the sample before being mixed was calculated in the way of deviation. The maximum deviation of the Tmax value of the DX8 well was 2 °C. The mud samples' Tmax values of the C58 well before being mixed were all distributed between 457 °C and 458 °C. The Tmax

210748914

values of the samples after being mixed were between 458 °C and 459 °C, with an average of 459 °C. Identically 459 ° C was used as the standard value of the mudstone sample, and the Tmax value of the sample before being mixed needed deviation calculation. The maximum deviation of the Tmax value of the C58 well was 2 °C. The rock pyrolysis experiment before and after being mixed of the samples show that the Tmax of the pyrolysis parameter before and after being mixed conforms to the relative double difference and deviation, which requires the GB/T 18602-2012 of rock pyrolysis analysis standard, with its deviation Should be ≤ 5 °C. That is to say, the current mixing method of the sample can fully meet the standard requirements (Table 1). Table1. The rock Pyrolysis data sheet before and after sample being mixed The effect of sample weight on Tmax Rock pyrolysis analysis sample size is generally specified as 100mg. If the sample amount is too small, it will affect the accuracy of the data. If the sample amount is too large, the pyrolysis hydrocarbon peak S2 will be too large, exceeding the detection range of the hydrogen flame ionization detector. The sample weight should be adjusted according to the organic matter abundance of the rock sample. The stable S2 and Tmax should be regarded as the true S2 and Tmax values within a certain range of sample weight. In this paper, seven mudstones, carbonaceous mudstones, coal and standard samples with different organic carbon contents were

210748914

selected. From 5mg or 10mg as the starting point, 100mg was sequentially determined, and the influence of the sampled amount on the pyrolysis parameters S2 and Tmax was analyzed ( Table 2). According to the experimental results, the S2 value is low when the sample amount is too small. For mudstone or tuff samples with S2 value less than 30mg/g, the S2 value does not vary with the sample volume. The sample size is greater than 40mg. For carbonaceous mudstone samples with S2 value greater than 30mg/g, the S2 value does not follow the sample weight. The variation range of the sample is from 20 mg to 35 mg. The range of S2 value of coal sample varies greatly. From the experimental data, for coal samples with S2 value less than 100mg/g, the suitable sample size ranges from 15mg to 30mg, for coal with S2 value greater than 100mg/g and less than 200mg/g. For example, the suitable sample size ranges from 5 mg to 15 mg. For coal samples of M006 well with S2 value greater than 200 mg/g, the sample weight exceeds 5 mg, and the S2 exceeds the detection range of the hydrogen flame ion detector. As the amount of sample increases, the value of S2 decreases. Therefore, a suitable amount of weighing is less than 5 mg. For Tmax, when the sample quantity does not exceed the S2 detection limit, the sample quantity has no effect on Tmax, but when the sample quantity is too much, it exceeds the detection limit of S2,

210748914

for example, the coal sample of M006 well increases with the sample quantity. The value of S2 decreases, and Tmax rises first and then decreases. Effect of soluble organic matter on Tmax Tmax is the peak temperature of the pyrolysis S2 peak, which is the temperature at which the hydrocarbon generation rate of the pyrolysis hydrocarbon is the largest. Therefore, Tmax is closely related to the pyrolysis peak. The incorporation of soluble organic matter increases the S2 value and affects the determination of the Tmax value. In this paper, the source rock samples of the source and reservoir in the Jimusaer area were selected, and the soluble organic matter in the rock was removed by chloroform extraction. The Tmax value of the rock pyrolysis parameters was compared before and after being extrated. The experimental results show that the soluble organic matter in the source rock will enter the S2 peak, resulting in a large S2 value and a decrease in the Tmax value. This is because under the conditions of conventional pyrolysis analysis, S2 is heavy at the initial analysis temperature of 300 °C. The soluble organic matter fails to fully evaporate, and some of them enter the pyrolysis hydrocarbon S2, while the Tmax of the soluble organic matter is low, so the Tmax value is lowered. It can be seen that when the chloroform pitch A is less than 1%, the Tmax value of samples varies by less than 2 °C, within the tolerance of the Tmax value. When the chloroform pitch A is greater

210748914

than 1%, the Tmax value ranges from 1 °C to 11 °C, the Tmax of most samples varies by more than 2 °C. Overall, the Tmax value of the sample after extraction is increased (Figure1 and Figure2). Therefore, when the chloroform pitch A is greater than 1%, it is necessary to extract the sample and then perform pyrolysis analysis to obtain an accurate Tmax value. Effect of minerals on Tmax Select the common clay minerals in the Junggar Basin, kaolinite, illite, and chlorite are mixed with type I, type II, and type III kerogens at a ratio of 1:20, 1:60, and 1:100. Pyrolysis experiments were carried out to analyze the effect of different clay minerals on the Tmax value of rock pyrolysis parameters. The experimental results show that, the three clay minerals have no effect on the pyrolysis Tmax value of type I kerogen, for type II kerogen, as the clay content increases, the Tmax value decreases, of which chlorite to Tmax The value of the value is greater than that of illite and kaolinite, and the type III kerogen increases with the increase of kaolinite content, and the Tmax value increases first and then decreases with the increase of illite and chlorite content (Figure3). Figure. 3 Comparison of Tmax form different clay minerals and different types of kerogen Conclusion (1) Before and after being mixed, the pyrolysis parameter Tmax meets the standard double difference and deviation requirement, and the mixed sample method has little effect on the result. (2) If the sample weight is too

210748914

small, the S2 value will be low. For mudstone, the appropriate sample weight is greater than 40mg, for carbonaceous mudstone samples with S2 value greater than 30mg/g, the sample weight is 20mg ~ 35mg, for coal samples with S2 value less than 100mg / g, the appropriate sample weight The range is from 15mg to 30mg. For coal samples with S2 value greater than 100mg/g and less than 200mg/g, the appropriate sample size ranges from 5mg to 15mg. For coal samples with S2 value greater than 200mg/g, the appropriate sample size range It is less than 5 mg. (3) Soluble organic matter leads to an increase in S2, and a decrease in Tmax. (4) Clay minerals have an effect on pyrolysis analysis of Tmax, but different types of clay minerals have different effects on different types of kerogen, and with the ratio of kerogen/clay minerals Change and the Tmax will change.

210748914

Immune Ecosystem of Virus-Infected Host Tissues Virus infected host cells serve as a central immune ecological niche during viral infection and replication and stimulate the host immune response via molecular signaling. The viral infection and multiplication process involves complex intracellular molecular interactions between viral components and the host factors. Various types of host cells are also involved to modulate immune factors in delicate and dynamic equilibrium to maintain a balanced immune ecosystem in an infected host tissue. Antiviral host arsenals are equipped to combat or eliminate viral invasion. However, viruses have evolved with strategies to counter against antiviral immunity or hijack cellular machinery to survive inside host tissue for their multiplication. However, host immune systems have also evolved to neutralize the infection; which, in turn, either clears the virus from the infected host or causes immune-mediated host tissue injury. A complex relationship between viral pathogenesis and host antiviral defense could define the immune ecosystem of virus-infected host tissues. Understanding of the molecular mechanism underlying this ecosystem would uncover strategies to modulate host immune function for antiviral therapeutics. This review presents past and present updates of immune-ecological components of virus infected host tissue and explains how viruses subvert the host immune surveillances. Introduction The immune ecosystem of a virus-infected host can be defined as the systemic interaction between the virus and the host immune system, resulting in either viral clearance or immune-mediated host tissue injury [1,2]. Remarkable havoc to people's health can stem from highly mutative viruses. Viruses are ever evolving to subvert the immune response, causing

13682346

emerging, adventitious, or even catastrophic diseases [3]. The immune system is continuously under viral assault, and to counter the invasion threat, it is well known that mammalians have developed a strong innate immune system and an intricate and specialized adaptive immune system to counteract the highly evolving virus infections. Immune system evolution occurs in all hosts from unicellular to vertebrates, for example, viral infection in protozoa leads to viral replication without apoptosis, while in multicellular organisms, apoptosis occurs during viral infection and vertebrate's response is more complex leading to long term immunity against the invading virus [4][5][6]. With the continuous efforts to uncover complex immune ecosystem mechanisms, scientists have innovated new tools to strengthen the immune system and neutralize the viral infections. For example, introduction of vaccines utilizes the ability of immune systems to produce powerful effective antibodies that can selectively neutralize immunogenic agents. Thus, vaccines enhance the host adaptive immunity and make the host able to counteract invasions of viruses to which it has never been naturally exposed before [1]. Better understanding of how the innate immune system works and how it evolves to fight the continuously mutating pathogens started with the pattern recognition hypothesis proposed in 1989 [7]. Briefly, pathogens carry molecular signatures called pathogen associate molecular patterns (PAMP), which can be recognized by pattern recognition receptors (PRRs) of the host cells. Upon interaction with PAMPs, PRRs become activated and trigger a cascade of immune responses against the viral infection. To date, a number of PPRs have been identified. For example, the toll-like receptor

13682346

(TLR) family can recognize a variety of different PAMPs [8]. Single-stranded RNA (ssRNA) can function as ligands for TLR7/8, thus sensing RNA virus infections. On the other hand, double-stranded RNA (dsRNA) serve as a ligand for TLR3, thus sensing the viral replication [9][10][11][12]. In addition, host cells are able to detect viral infection through TLR-independent signaling pathways involving other cytoplasmic RNA helicase proteins, such as retinoic acid inducible gene (RIG-I) and melanoma differentiation-associated gene 5 (MDA5), etc. [13]. Immune sensing through innate signaling of virus-infected cells eventually induces the translation of interferon (IFN), these secreted IFNs bind with their cognate receptors present on the cell membrane. This binding activates a signaling pathway leading to expression of various IFN-dependent antiviral molecules [14]. Furthermore, adaptive immunity carries on the duty of viral clearance in the latter stages. In spite of the existence of several correlative host defense lines to neutralize invasive pathogens or restrict the viral life cycle, viruses have evolved and adapted diverse counter strategies to evade the host immunity, sustaining continuous replication and endurable infection in the host using avoidance and escape tactics. Some viruses have developed highly complicated mechanisms such as viral growth in immunologically privileged sites, e.g., herpes virus establishes a latency in sensory neurons to avoid immune surveillance [15]; antigenic drift, e.g., influenza virus uses this to evade the B-cell immunity [16,17]; induced expression of some factors against innate immunity, e.g., human immune deficiency virus (HIV) can block the interferon induction in dendritic cell using its proteins Vpr and Vif [18][19][20][21][22][23][24]; and other

13682346

more diverse and intricate maneuvers. In this review, we make a comprehensive overview of virus-host interactions, especially immune ecosystem of virus-infected host. We also highlight the virus subversion mechanisms against host immune responses. Immune Sensing of Viral Infection Host immune response is first activated by sensing the viral infection through PRRs [25]. The host cells interact with the virus in several different viral states, including extracellular native virus, intracellular viral components, and the viral replicate intermediate. Immune sensing of viral infection is mainly achieved by specialized cells and cellular factors to detect specific viral elements in different viral forms [26]. The immune response differs according to type of virus and route of infection [27]. For instance, a study showed that upon infection with inactivated whole influenza virus vaccine containing viral ssRNA, the TLR7/myeloid differentiation primary response 88 (MYD88) pathway was the only pathway triggered without any activation of the RIG-I/IPS-1 pathway, though both of them are parallel innate immune pathways [28]. For DNA viruses, TLR9 can recognize the unmethylated DNA and share the downstream signaling pathway with the adapter protein MYD88 [29]. Viral DNA is also recognized by cGAS in the cytoplasm and IFI16 in the nucleus. Both cGAS and IFI16 recruit a common adaptor protein, Stimulator of interferon genes (STING), which signals to TBK1 for the activation of IRF3, which ultimately induces the expression of type I IFN and other antiviral genes [30]. However, the innate immune response against a DNA vaccine containing CpG-DNA was triggered only by the TBK1 not the TLR9/MYD88 pathway [31].

13682346

Intercellular Immune Ecosystem of Virus-Infected Tissues In multicellular organisms, cells are actively working as a single fundamental unit countering the viral infection. Once the virus attaches to host cells, the host cells starts a series of events to alert the neighboring cells against the invader and trigger the effector cells and pro-inflammatory response. This happens via the production of cytokines which help the neighbor cells to produce some inhibitory effects on viral infection and replication, e.g., IFNs; a potent neutrophil chemoattractant, e.g., CXCL8; other cell chemoattractants (monocytes, eosinophils, and T cells) e.g., CCL2, CCL3, CCL4, and CCL5; or other cytokines that lead to the acute-phase viral removal, e.g., interleukin (IL)-6 [32,33]. Knowing that the pathogen replication speed is a cornerstone in the viral pathogenesis, establishment of such an intercellular immune ecosystem including intercellular interaction and intracellular signaling is considered beneficial to host antiviral defense, which is able to clear the pathogen and limits its spread in infected tissues without waiting for the classical immune response [34]. Not only viruses are undergoing modifications and development, the tissue's microenvironment also undergoes several modifications after a successful infection resolution. These modifications can be in favor of the next infection or to counter it. For example, after severe lung infection, severe lung tissue damage leads to a repair process that changes the lung matrix composition, (such as more collagen and fibronectin deposition) providing additional binding sites for bacteria [35]. Successive lung infections can change the lymphatic network and the frequency of inducible bronchus-associated lymphoid tissue (iBALT) [36,37]. In addition, specific

13682346

memory T cells persist at the infection site and are termed resident memory T cells. These cells are resident within the infection tissues and can promote the early innate immune activation for the recurrent infection. Skin resident CD49a + cells were reported to express perforin and granzyme B after treatment with IL-15 [38][39][40]. It is worth mentioning that natural killer (NK) cells are critical for early non-specific resistance against the viral infection [41]. NK cells are specialized lymphocytes lacking antigen-specific receptors, yet they are able to demolish tumor cells, virus-infected cells, and any cell in the state of stress [42]. NK cells are able to differentiate between normal healthy cells and abnormal cells via certain sophisticated attributes of the cellular surface receptors. The leading receptor for NK cells is the class I major histocompatibility complex (MHC); unstable expression of the class I MHC means an unbalanced cellular state resulting via serial of cascades in the activation of NK cells [43]. However, receptors such as NKG2D and NKp30 are known to help the NK cells to differentiate between healthy and unhealthy cells [44,45]. NK cells have a higher tendency to lyse cells lacking surface class I MHC expression. NK cells lyse the virus-infected cells with the help of cytotoxic T lymphocytes (CTL), inducing cellular apoptosis using its cytolytic granules (containing perforin, granzyme A, and granzyme B) [46]. In the host-virus ecosystem, the virus has evolved counter defenses to help in the continuity of the infection cycle. Some viruses have developed strategies to delude the NK cells or

13682346

disrupt the class I MHC antigen presentation, thus curbing the NK cells. For example, poliovirus protein 3A interacts with the endoplasmic reticulum (ER) membrane to cease protein transport from the ER to the Golgi apparatus, hence preventing the transport of the MHC-bearing polio-specific peptide to the cell membrane [47]. Foot and mouth disease (FMD) inhibits protein transport using its viral 2BC protein [48]. The Tat protein of retroviruses interferes with class I MHC messenger RNA (mRNA) transcription [49]. Relocalization of the class I MHC to the trans-Golgi network by the retrovirus Nef protein results in the downregulation of the surface expression of MHC-I [50]. Cytomegaloviruses (CMV) can resist the NK cells attack though the severe down-regulation of the class I MHC expression by down-regulation of important proteins (e.g., UL18) that are required for the NK cell stimulation [51]. Primarily, the HLA-C and HLA-E molecules protect normal cells from the NK mediated cell lysis. It was believed that retroviruses like HIV-1 can selectively disrupt the expression of HLA-A and HLA-B but not HLA-C and HLA-E, thus, the infected cells are less likely to be killed by NK cells [52]. Yet, recent research found that HLA-C is downregulated by most primary HIV-1 clones. The viral Vpu protein was reported in this study to reduce the ability of HLA-C restricted CTLs to suppress viral replication in CD4 + cells in vitro [53]. Herpesviruses possess a special strategy to coexist with the immune system by encoding several genes that interfere with the MHC-I antigen presentation [54][55][56]. Alphaherpesviruses exploit the fact

13682346

that neurons are immunologically privileged and have a lower expression of class I MHC compared to other cells; they start a long life latent infection in neurons and express no protein during their latency, thus escaping the CTL immune surveillance [57]. Epstein-Barr virus does express a latency protein but with a glycine-alanine repeats domains that bind to proteasomes, making the latency protein undetectable by class I MHC [58]. It was believed that only the B and T lymphocytes of the adaptive immune response can possess memory that is able to recognize a repeated infection, until several studies challenged this hypothesis [59]. Studies introduced the new term "trained immunity" that describes the enhanced immune response following previous exposure to some immunogenic agents leading to a robust response against related or unrelated pathogens in the recurrent infection [60,61]. Some studies suggested that NK cells could keep a memory of the previous antigens to make them able to mediate a more robust immune response [62][63][64]. Also, monocytes have been reported to possess the same phenomena [65]. Intracellular Immune Ecosystem of Virus-Infected Cells As viruses are obligate intracellular parasites, infected host tissue serves as a central immune ecological niche during viral genome transcription, replication, and stimulation of the host immune response via molecular signaling. The crucial step in the process of the viral invasion is the attachment, which eventually leads to viral recognition by the immune system. Several endocytic pathways are involved in the interaction with the infectious viral components. As a result, the endosomal sensors are an important spot

13682346

for innate immunity ( Figure 1). The immune response can be activated either by detection of the viral PAMPs or the immune and inflammatory cytokines. Based on the nature of nucleic acid, viruses are classified into DNA and RNA viruses. The majority of viruses infecting animals are RNA viruses. All RNA virus replication proceeds through a RNA strand complimentary intermediate, except for in retroviruses where the intermediate is DNA [66]. Despite differences in genomic features and replication strategies, immediately after virus infection, all RNA viruses trigger evolutionarily conserved innate immune responses that serve as a first line of defense against infection. Immune receptor PRRs, comprised of the key families TLRs, RIG-I like receptors, nucleotide-binding oligomerization (NOD) like receptor, and others, recognize specific PAMPs and thereby stimulate multiple signaling molecular cascades and induce transcription of nonspecific immune effector genes [67]. It is important to know that sensing of the DNA viruses needs a different set of receptors and immune signaling pathways that have already been reviewed in detail (see review in [68]). The difference between the immune sensing and response between the RNA and DNA viruses was also reported and reviewed in detail (see review in [68,69]). Additionally, non-coding RNAs (ncRNAs) have recently gained wide research interest. In addition, the key roles of ncRNAs in the immune response against viral infections has been established and reviewed already (see review in [12,13,[70][71][72]). Importantly, sensing of PAMPs by PPRs remarkably up-regulates the genes involved in the inflammatory response encoding pro-inflammatory cytokines/chemokines and interferons (IFNs) that induce antiviral gene products

13682346

(Figure 1) [73]. Production of type I IFNs plays an important role in the induction of antiviral responses, which triggers transcription of IFN-inducible antiviral genes (ISGs) (Figure 1). On the other hand, viruses have evolved several strategies to hijack the host cellular machinery and, then, shut off the host cell gene expression at both the transcriptional and translational level. For example, the influenza A virus (IAV) has been reported to induce the degradation of eukaryotic translation initiation factor 4B [74]. Some viruses encode proteins to protect viral nucleic acid from being detected by cytoplasmic sensors. PRR cGAS can sense the HIV viral complementary DNA (cDNA) in the cytoplasm; HIV-1 but not HIV-2 cDNA is protected within the viral capsid until it is translocated to the nucleus for replication. That is due to the affinity of the HIV-1 capsid for stabilization by the host protein cycophilin A (CypA), preventing its exposure to the cGAS in the cytoplasm [75]. Moreover, viruses use a variety of strategies to subvert the interferon response, and these are discussed later in this review [76]. affinity of the HIV-1 capsid for stabilization by the host protein cycophilin A (CypA), preventing its exposure to the cGAS in the cytoplasm [75]. Moreover, viruses use a variety of strategies to subvert the interferon response, and these are discussed later in this review [76]. Although TLRs are the primary host defense sensors to combat viral invasion, viruses have evolved to subvert TLR-medicated antiviral immunity. Xagorari and Chlichlia [9] reviewed the perturbation of TLR-medicated immunity by numerous viruses

13682346

[9]. For example, the P protein of the measles virus suppresses TLR signaling through up-regulation of the ubiquitin modifying enzyme A20 [83]. Hepatitis C virus (HCV) inhibits activation of NF-κB and IRF3 by proteolysis of TRIF (the adaptor protein, which links TLR3 and kinase). HCV utilizes the viral NS3/4A to cleave TRIF, which is an intermediate in TLR3, mitochondrial antiviral-signaling protein (MAVS), and RIG-I signaling pathways, therefore, dsRNA cannot induce the IFN production through these pathways [84][85][86]. Recently, an Orf virus (ORFV) virion-associated protein, ORFV119, was identified that inhibits the NF-κB signaling very early in infection [87]. As described above, TLR are mostly endocytic viral nucleic acid sensors [88]. Yet, TLR2 and TLR4 are classical microbial PRRs that recognize microbial moieties and are examples of the surface receptors involved in the recognition of the viral proteins and trigger the pro-inflammatory responses [89]. For example, TLR4 was reported to be triggered by infection with retrovirus, respiratory syncytial virus, and mouse mammary tumor virus [78,90]. TLR2 can detect the hepatitis C virus (HCV), herpes simplex virus (HSV), human cytomegalovirus, and measles particles [91][92][93]. PRRs-dependent activation of signaling pathways is critical for the intracellular immune ecosystem of virus-infected cells. RIG-I Like Receptors RIG-I like receptors (RLRs) are cytosolic intracellular key receptors, classically responsible for sensing non-self RNA signatures. Yet, several studies suggested that RIG-I is required to sense the DNA viruses for the induction of type I IFN [94][95][96]. The RLR family, comprised of RIG-I, MDA5, laboratory of genetics and physiology 2, and a homolog of mouse D11lgp2 (Laboratory

13682346

of Genetics and Physiology 2 (LGP2)), expressed in most of the tissue types (see review in [84]). RLRs contain a central DExD/H-box helicase domain and a C-terminal domain responsible for binding viral RNA. RIG-I and MDA5 share structural similarities. They have two N-terminal caspase activation and recruitment domains (CARDs) for downstream signaling. LGP2 lacks the CARDs and is thought to play a regulatory role in RLR signaling [97]. RIG-I plays an important role in the detection of orthomyxoviruses, rhabdoviruses, and arenaviruses, and MDA5 preferentially detects picornaviruses. Additionally, many other viruses such as flaviviruses, paramyxoviruses, reoviruses, and others are sensed by both RIG-I and MDA5 [98]. Upon detecting the viral RNA ligands in the cytoplasm, RLRs trigger innate immunity and inflammation. The role of RLRs mediated immune signaling in influenza virus infection has been already reviewed with highlights of the important downstream key immune components involved in the activation of IRFs and NF-κ-B [13,99]. An important feature of RIG-I and MDA5 mediated non-self RNA sensing is the activation of the transcriptional factors such as IRF3/7, NF-κB, and Activating transcription factor-2 (ATF2)/c-Jun to induce the transcription of IFNs and pro-inflammatory cytokines. Activation of these transcriptional factors eventually induces the expression of hundreds of ISGs [99]. Viruses show a variety of strategies to evade RLR-mediated immune responses. Importantly, the NS1 protein of the influenza A virus inhibits the functional RIG-I and RIG-I dependent activation of NF-κB [100]. Z proteins of pathogenic arenaviruses showed the ability to interact with RIG-I and MDA5, causing the inhibition of the latter two and

13682346

leading to a significant inhibition of type I interferon (IFN) responses [101]. The non virion (NV) protein of fish Novirhabdovirus showed the ability to counteract RIG-I and TBK1-dependent interferon and IFN-stimulated gene promoter induction in fish cells resulting in the suppression of the anti-viral state induction [102]. V proteins of paramyxoviruses interact or inhibit MDA5 and LGP2 [103,104]. HCV protease NS3/4A cleaves MAVS at Cys-508 resulting in the dislocation of the N-terminal fragment of MAVS [105]. Some positive-sense RNA viruses like porcine reproductive and respiratory syndrome virus (PRRSV) can cleave the MAVS during infection using their nsp4 cysteine protease [106,107]. Recently, the MDA5-Mediated innate immune response has been reported to be disrupted by different proteins encoded by some picornaviridae viruses, resulting in the inhibition of the viruses interaction with MAVS, which was followed by inhibition of the MDA5-dependent translation of type I IFN [108]. NOD-Like Receptors (NLRs) NLRs are cytosolic PRRs. They detect diverse PAMPs or damage associated molecular patters (DAMPs) produced by viral infection. The human genome encodes 22 NLRs. Of them, nucleotide-binding oligomerization 2 (NOD2) and multi-protein inflammasome complex (i.e., NOD-, LRR-and pyrine domain-containing 3 (NLRP3)) are well studied for sensing the viral infections. NOD2 was reported to induce IFNs production following ssRNA transfection and respiratory syncytial and influenza A virus infection, whereas NLRP3 inflammasome activation was reported in viral infections such as IAV, encephalomyocarditis virus, and HCV [109,110]. Regulation and activation NLRP3 inflammasomes by viroporins of animal viruses have been already reviewed (see review in [111]). NLRP3 inflammasomes are activated by damage, inflammation,

13682346

or stress, and this activation leads to the production of active IL-1β and IL-18 by activating caspase I in IAV infection, while NLRP3 can activate the production of cytokines by triggering signal 1 and signal 2 during infection with IAV [13]. The V protein of measles and the NS1 protein in influenza A virus are representative examples of inhibition of the NLRP3 inflammasome [112,113]. Interferons Function as Critical Components in the Immune Ecosystem of a Virus Infected Tissue Interferons (type I to III) are critical immune-ecological components produced by virus-infected cells; they function as antiviral molecules and immune-modulators. Roles of type I IFNs (IFN-α and IFN-β) are well characterized and intensively explored. IFNs bind to their respective receptors to activate the Janus kinase-signal transducer and activator of transcription (JAK-STAT) pathway that governs antiviral defense [114]. IFNs-activated immunity determines the extent of host susceptibility towards viral infection. Recently, mice lacking interferon α receptor (IFNAR) (type I IFN receptor) were reported to be highly susceptible to pseudorabies virus infection [115]. Remarkably, binding of type I and type III IFNs to their respective receptors activates tyrosine kinase 2 (TYK2) and Janus kinase 1 (JAK1). This activation leads to the recruitment and phosphorylation of signal transducers and activators of transcription, signal transducer and activator of transcription 1 (STAT1) and STAT2. Phosphorylated STAT 1 and STAT 2 hetero-dimerize and are assembled with IFN-regulatory factor 9 (IRF9) to form a tri-molecular complex called IFN-stimulated gene factor 3 (ISGF3). The complex, then, translocates into the nucleus. Inside the nucleus, ISGF3 binds to its

13682346

cognate DNA sequences, which are known as IFN-stimulated response elements (ISREs; consensus sequence TTTCNNTTTC), thereby, directly activating the transcription of ISGs (see review in [116,117]). On other hand, type II IFN binds to its receptor (IFN-γ receptors 1 and 2 heterodimers) and leads to the formation of phosphorylated STAT1 (pSTAT1) homo-dimers. Phosphorylated STAT1 homo-dimers form the IFN-γ activation factor (GAF). Following GAF nuclear translocation, GAF binds to the gamma-activated sequence (GAS, TTCNNNGAA) in the promoter region of the ISGs, and this results in the expression of antiviral ISGs [116,117]. Furthermore, IFN-γ plays a pivotal role in regulating the immune function and bridging the innate and adaptive immune responses [118]. Importantly, non-canonical/alternative pathways have been emerging to uncover hidden components of the immune-ecosystem. For example, an alternative STAT signaling pathway acts in antiviral immunity in Caenorhabditis elegans [119]. Activated JAK1 also activates other members of the STAT family (STAT1-6) [120] and induces several alternative signaling pathways [13,76,104,109,121]. Activation of all pathways, eventually, further amplifies the amplitude of IFN production and signaling. Networks of cellular pathways that regard IFNs as critical immune components in virus infection have been extensively reviewed elsewhere (see review in [72,122]). IFN-mediated pathways obviously stimulate expression of hundreds of ISGs by activating the specific transcriptional regulator factors. Representative ISGs are MxA, protein kinase R (PKR), 2 -5 Oligoadenylate Synthetase (OAS), ISG15, viperin, tetherin, IFTIMs, RIPK2, IFI16, and so on [14,32,70,[123][124][125]. Recently, many more functional ISGs have been identified through protein-wise or genome-wise screening. Nevertheless, arguably, intensive studies are required to define the roles of

13682346

these ISGs in the immune ecosystem of a virus infected host. Although IFNs act as critical immune components, still, viruses can perturb the IFN-mediated immune barriers. The balance of immune ecosystems is always unsteady. Supportive and virtual host immunity is always perturbed by certain types of viruses [76]. Recently, tetherin has been shown to inhibit type I IFN via targeting MAVS [18]. A viral-hijacked E3 ubiquitin ligase is also shown to shut off IFN signaling [33,126]. A variety of IFNs subversion strategies by viruses have been extensively reported over the past 10 years [76]. Importantly, numerous viruses exploit suppressors of cytokine signaling (SOCSs) to inhibit IFN signaling [34,127]. The suppression of type III IFN signaling by virus-induced SOCS-1 was reported to cause an adaptive increase in type III IFN expression by the host to protect cells against the viral infection, as a consequence, it lead to excessive production of the IFNs with impaired antiviral response [118]. Some viruses can encode antagonists for type I IFN, such as the vaccinia virus that embraces the C9 ankyrin repeat/F-Box protein that has been recently considered as an antagonist of the Type I IFN-induced antiviral state [128]. Kaposi's sarcoma-associated herpesvirus (KSHV) can selectively interact with IRF7, inhibiting the IRF7 dimerization and leading to the suppression of IRF7-mediated activation of type I IFN [129]. The pestiviral protein N (Npro) was reported to interact with the IFN regulatory factor 3 (IRF3) via binding to the active form of IRF3 in the presence of its transcriptional coactivator, CREB-binding protein (CBP), resulting in the

13682346

inhibition of the activation of type I alpha/beta IFN [130]. DNA viruses such as adenovirus, human papilloma virus, and Kaposi's sarcoma virus utilize their viral proteins to bind with the DNA adaptor STING to prevent induction of type I IFN [131,132]. Viruses have also evolved to inhibit the action of ISGs. For example, protein kinase R (PKR) is one of the antiviral effector ISGs; human cytomegalovirus virus encodes pTRS1 and pIRS1 proteins that antagonize PKR, preventing its autophosphorylation of eukaryotic initiation factor 2 alpha (eIF2α) and facilitating the viral replication [133]. Vault RNAs (vtRNAs) have been reported to promote IAV replication via the inhibition of PKR activation and the subsequent IFN response [125]. Taken all together, virus-induced IFNs production and suppression of their signaling seem to be in the state of delicate and dynamic equilibrium by modulating host factors to maintain a balanced immune ecosystem [12]. Destruction of such a balanced immune ecosystem is a primary cause of viral pathogenesis. Adaptive Immune Response to Viral Infection The primary adaptive immune response usually takes several days and starts with the binding of an antigen with its specific receptor on the T cells or B cells. Migration of the pathogen stimulates dendritic cells of the draining lymph nodes. This is followed by several steps that ends with the release of lymphocytes embracing the antigen-specific receptors and the production of effector and memory cells. The adaptive immune response to viral infection has two-main arms: humoral immunity and cellular immunity. Humoral immunity consists of the antibodies secreted by plasma cells

13682346

that can neutralize the native extracellular viruses. The cellular immunity is driven by α-β T cell receptors expressed by T lymphocytes that recognize the antigen processed peptide bound to MHC molecules on the surface of infected cells [134,135]. After the primary exposure to a certain pathogen, adaptive immune response confers long-term, often lifelong, protection against the exposed pathogen. This is due to the fact that adaptive immune response exhibits memory T cells or B cells to provide anearly response against the recurrent infection. In case of reinfection, memory B cells are responsible for the generation of an accelerated and more robust antibody-mediated immune response [136]. Memory T cells are the T lymphocytes that were previously exposed to the antigen, either via natural or artificial exposure, so at the second encounter, the memory T cells can reproduce quickly to mount a faster and stronger immune response than was seen during the first infection [137]. Memory T cells are sub-classified into two important types: central memory T cells (T CM ) and effector memory T cells (T EM ). T CM is capable of the production of interleukin (IL)-2 and reproduces extensively, while T EM is capable for the production of effector cytokines like IFN-γ [137]. This mechanism allowed great progress in the production of vaccines to deleterious viral infections that impact the economy and public health. Exploiting the adaptive immune response and the memory cells produced by the immune system, we are now able to protect humans, animals, and plants from diseases that they have never been

13682346

exposed to, via exposing them to an artificial compound containing either conventionally inactivated, live-attenuated virus vaccines, recombinant viruses that express protective proteins of heterologous viruses, virus-like particles (VLPs), or DNA vaccines [134,138]. Various viruses have evolved to invade the effector cells to evade the adaptive immunity. For example, HIV can infect the CD4 + T H cells, resulting in a serious immune suppression due to cellular lysis [139]. Measles virus infects B cells, CD4 + , and CD8 + memory T cells and monocytes, resulting in immune suppression that lasts for several weeks after the virus invasion [140]. Epstein-Barr virus infects the B cells, causing impairment of the antigen recognition and antibody release [141]. In addition, some viruses can infect the thymus in the early animal life, leading to its identification as a non-immunogenic antigen and, therefore giving immune tolerance as the virus is no longer recognized as a foreign antigen. This results in a long-life infection of the animal. Examples of these viruses are lymphocytic choriomeningitis virus [142] and murine leukemia virus [143]. On the other hand, hepatitis B virus (HBV) infection has a unique infection style in infants or young children via causing an asymptomatic disease phase (immune tolerant phase) characterized by high HBV titers and a low incidence of liver inflammation and immune response to the virus [144,145]. The mechanisms underlying the immune tolerance is still unclear, however, they might be due to ineffective antigen processing and transport to major histocompatibility complex class I molecules [146] leading to HBV-specific T-cell hyporesponsiveness [147]. HBV

13682346

can also harness the young children's developing immune system and make the fetal immunity facilitate its persistence in patients after the prenatal exposure, causing a persistent long-life infection [148]. Many other viruses have developed persistence mechanisms to cause permanent infection and to persist indefinitely within the host [149]. Herpes virus can cause nonproductive infection by developing the herpes virus latency in immune privileged sites [150]; retreoviruses cause their persistent infection via integration of their provirus into the host genome [151]; and some other viruses cause a persist infection through continuous viral replication, e.g., filoviruses [152], arenaviruses [153], and polyomaviruses [154,155]. Conclusions Dynamic interaction between the virus and host immune system results in the formation of a complex immune ecosystem involving intermolecular signaling webs for their synergistic or antagonistic fitness and survival. The host cell modulates intracellular components to clear the virus away, but viruses hijack host cell components, simultaneously. To guarantee successful and durable defense mechanisms against virus infection, hosts have evolved a highly intricate, sophisticated, and adaptable immune system to protect against continuously emerging threats and mutated viruses. Antiviral immune responses comprise complex networks of innate and acquired defenders, some of them are specific to certain pathogens and some of them are nonspecific to counteract any state of stress or immunogenic particle. It is well known that innate immunity provides the rapid nonspecific response, while adaptive immunity is only developed after the initial virus exposure. The interplay between the innate and adaptive immunity will provide the desirable immune ecosystem that can challenge the virus infection

13682346

and demolish the infected cells. On the other hand, viruses also evolve. Viruses have co-evolved with their hosts to produce remarkable strategies to counteract the host defenses. These strategies include the rapid shutdown of host molecular synthesis, evasive strategies of viral antigen production, interference with MHC class I and class II antigen presentation, impairing NK cell function, suppression of antiviral cytokine signaling, and blocking apoptosis. It is recognized that health can be compromised in are remarkable way by highly mutative viruses. Viruses are ever evolving and being transmitted to susceptible hosts from animal or environmental sources to cause diseases, such as often catastrophic emerging and adventitious diseases. Virus fitness inside a host is critically reliant on the host immune ecosystem and its strategies to counter the immune ecosystem. Approaches in the understanding and management of host immune ecosystem and virus-host community would further develop novel antiviral therapeutics. Conflicts of Interest: The authors declare no conflict of interest.

13682346

High resolution optical and near-IR imaging of the quadruple quasar RX J0911.4+0551 We report the detection of four images in the recently discovered lensed QSO RX J0911.4+0551. With a maximum angular separation of 3.1", it is the quadruply imaged QSO with the widest known angular separation. Raw and deconvolved data reveal an elongated lens galaxy. The observed reddening in at least two of the four QSO images suggests differential extinction by this lensing galaxy. We show that both an ellipticity of the galaxy (epsilon_{min}=0.075) and an external shear (gamma_{min}=0.15) from a nearby mass has to be included in the lensing potential in order to reproduce the complex geometry observed in RX J0911.4+0551. A possible galaxy cluster is detected about 38", from RX J0911.4+0551 and could contribute to the X-ray emission observed by ROSAT in this field. The color of these galaxies indicates a plausible redshift in the range of 0.6-0.8. Introduction RX J0911.4+0551, an AGN candidate selected from the ROSAT All-Sky Survey (RASS) (Bade et al. 1995, Hagen et al. 1995, has recently been classified by Bade et al. (1997; hereafter B97) as a new multiply imaged QSO. B97 show that it consists of at least three objects: two barely resolved components and a third fainter one located 3.1 ′′ away from the other two. They also show that the spectrum of this third fainter component is similar to the combined spectrum of the two bright components. The lensed source is a radio quiet QSO at z = 2.8. Since RASS detections of distant radio quiet

17220528

QSOs are rare, B97 pointed out that the observed X-ray flux might originate from a galaxy cluster at z ≥ 0.5 within the ROSAT error box. We present here new optical and near-IR high-resolution images of RX J0911.4+0551 obtained with the 2.56m Nordic Optical Telescope (NOT) and the ESO 3.5m New Technology Telescope (NTT). Careful deconvolution of the data allows us to clearly resolve the object into four QSO components and a lensing galaxy. In addition, a candidate galaxy cluster is detected in the vicinity of the four QSO images. We estimate its redshift from the photometric analysis of its member galaxies. Observations and reductions We first observed RX J0911.4+0551 in the K-band with IRAC 2b on the ESO/MPI 2.2m telescope on November 12, 1997. In spite of the poor seeing conditions (∼ 1.3 ′′ ), preliminary deconvolution of the data made it possible to suspect the quadruple nature of this object. Much better optical observations were obtained at the NOT (La Palma, Canary Islands, Spain). Three 300s exposures through the I filter, with a seeing of ∼ 0. ′′ 8 were obtained with ALFOSC under photometric conditions on November 16, 1997. Under non-photometric, but excellent seeing conditions (∼ 0. ′′ 5 − 0. ′′ 8), three 300s I-band exposures, three 300s V -band and five 600s U -band exposures were taken with HIRAC on the night of December 3, 1997. The pixel scales for HIRAC and ALFOSC are 0. ′′ 1071 and 0. ′′ 186, respectively. RX J0911.4+0551 was also the first gravitational lens to

17220528

be observed with the new wide field near-IR instrument SOFI, mounted on the ESO 3.5m NTT. Excellent K and J images were taken on December 15, 1997, andJanuary 19, 1998 respectively. The 1024 × 1024 Rockwell detector was used with a pixel scale of 0. ′′ 144. The optical data were bias subtracted and flat-field corrected using sky-flats. Fringe-correction was also applied to the I-band data from ALFOSC. Sky subtraction was carried out by fitting low-order polynomial surfaces to selected areas of the frames. Cosmic ray removal was finally performed on the data. The infrared data were processed as explained in Courbin, Lidman & Magain (1998), but in a much more efficient way for SOFI than for IRAC 2b data, since the array used with the former instrument is cosmetically superior to the array used with the latter. Deconvolution of RX J0911.4+0551 All images were deconvolved using the new deconvolution algorithm by Magain, Courbin & Sohy (1998;hereafter MCS). The sampling of the images was improved in the deconvolution process, i.e. the adopted pixel size in the deconvolved image is half the pixel size of the original frames. The final resolution adopted in each band was chosen according to the signal-to-noise (S/N) ratio of the data, the final resolution improving with the S/N. Our NOT HIRAC data were deconvolved down to the best resolution achievable with the adopted sampling (2 pixels FWHM, i.e. 0. ′′ 1), whereas the NOT/ALFOSC frames were deconvolved to a resolution of 3 pixels FWHM, i.e. 0. ′′ 29. Although of very good

17220528

quality, the near IR-data have a lower S/N than the optical data. The resolution was therefore limited to 5 pixels FWHM in both J and K (0. ′′ 36) in order to avoid noise enhancement. The MCS algorithm can be used to deconvolve simultaneously a stack of individual images or to deconvolve a single stacked image. The results from the simultaneous deconvolutions are displayed in Fig. 1. The four QSO images are labeled A1, A2, A3, and B. The quality of the results was checked from the residual maps, as explained in Courbin et al. (1998). The deconvolution procedure decomposes the images into a number of Gaussian point sources, for which the program returns the positions and intensities, plus a deconvolved numerical background. In the present case, the deconvolution was also performed with an analytical De Vaucouleurs, and exponential disk galaxy profile at the position of the lensing galaxy, in order to better describe its morphology. Table 1 lists the flux of each QSO component, relative to A1, as derived from the simultaneous deconvolutions. Although the HIRAC U , V and I band data were taken during non photometric conditions, they can still be used to determine the relative fluxes between the four images of the QSO. The errors in the relative fluxes are determined from the simultaneous deconvolutions and represent the 1-σ standard deviation in the peak intensities. Results When flux calibration was possible, magnitudes were calculated and the results are displayed in Table 2, which also contains the positions of the four QSO components

17220528

relative to A1. The astrometric errors are derived by comparing the positions of the components in the different bands. The deconvolved numerical background is used to determine the galaxy position from the first order moment of the light distribution. A reasonable estimate of the error on the lens position was derived from moment measurements through apertures of varying size and on several images obtained by running deconvolutions with different initial conditions. I, J, and K magnitudes for the galaxy were also estimated from the deconvolved background image by aperture photometry (∼ 1.2 ′′ aperture in diameter). The numerical galaxy is elongated in all three bands and the position angle of its major axis is θ G ≃ 140 ± 5 • . In the near-IR, the galaxy looks like it is composed of a bright sharp nucleus plus a diffuse elongated disk. However we can not exclude that the observed elongation is due to an unresolved blend of two or more intervening objects. None of our two analytical profiles fit perfectly the galaxy. The De Vaucouleurs profile (e = 1 − b/a = 0.31 ± 0.07, θ G = 130 ± 10 • ) fits slightly better than the exponential disk light distribution, but still produces residual maps with values as large as 1.5-2 per pixel, compared with a χ 2 of 2.5-5 per pixel for the exponential disk profile. The ellipticity and the position angle derived this way are very uncertain due to the low S/N of the data. Much deeper observations will be required

17220528

to perform precise surface photometry of the lens(es) and to draw a definite conclusion about its (their) morphology. Field photometry In order to detect any intervening galaxy cluster which might be involved in the overall lensing potential and contributing to the X-ray emission observed by ROSAT, we performed I, J, and K band photometry on all the galaxies in a 2.5 ′ field around the lensed QSO. A composite color image was also constructed from the frames taken through these 3 filters, in order to directly visualize any group of galaxies with similar colors, and therefore likely to be at the same redshift. The color composite is presented in Fig. 2. Aperture photometry was carried out using the SExtractor package (LINUX version 1.2b5, Bertin & Arnouts, 1996). The faintest objects were selected to have at least 5 adjacent pixels above 1.2σ sky , leading to the limiting magnitudes 23.8, 21.6, and 20.0 mag/arcsec 2 in the I, J, and K bands, respectively. The faintest extended object measured in the different bands had magnitudes 23.0, 22.0, and 20.3 in I, J, and K, respectively. The color-magnitude diagram of the field was constructed from the I and K band data which give the widest wavelength range possible with our photometric data. Since the seeing was different in the two bands, particular attention was paid to the choice of the isophotal apertures fitted to the galaxies. They were chosen to be as large as possible, still avoiding too much contamination from the sky noise, as an oversized isophotal aperture

17220528

would introduce. The color-magnitude diagram of the galaxies in the 2.5 ′ field is displayed in Fig. 3 . Stars are not included in this plot. Note that, because of their proximity, the magnitudes of the two blended members of the cluster candidate (center of the circle in Fig. 2) might be underestimated by as much as 0.3-0.4 magnitudes in both I and K. Models In a first attempt to model the system we chose an elliptical potential of the form where the coordinates x ′ and y ′ are measured along the principal axes of the galaxy, whose position angle θ G is a free parameter. For small ellipticities ǫ, this potential is a good approximation for elliptical mass distributions (Kassiola & Kovner 1993). Additionally, an external shear γ with direction φ is included. Even without assuming an explicit potential ψ, we can determine the minimal γ and ǫ needed to reproduce the observed image positions, by applying the methods given by Witt & Mao (1997). Their methods eliminate the unknown parameters of the model to find constraints for the ellipticity of the galaxy and the external shear. For the shear we predict a minimum of γ min = 0.15 ± 0.07, while ǫ min = 0.075 ± 0.034 (both with 1σ errors). Therefore, neither ellipticity nor shear should be omitted from the modeling. To keep models simple, we used a pseudo-isothermal potential corresponding to an apparent deflection angle of which degenerates to a singular isothermal sphere for vanishing core-radius r c and ellipticity ǫ.

17220528