Animal Model Studies Reveal that Common Human-Centric Non-Coding Variants from Epidemiology are By-products of Primate Evolution Unrelated to Physiological Control of Blood Pressure

Animal Model Studies Reveal that Common Human-Centric Non-Coding Variants from Epidemiology are By-products of Primate Evolution Unrelated to Physiological Control of Blood Pressure. Abstract Background : Human genome-wide association studies (GWAS) on blood pressure (BP) have been undertaken by avoiding its physiology and mechanisms controlling BP. Consequently, the physiological significance of GWAS on BP remains undiscovered. A shared mechanistic foundation starts to untangle human physiological regulations of BP as primate versions of rodents and vice versa . Thus, understanding mechanisms in rodents is equivalent to unraveling the same in humans rooted in their common ancestors. Methods : We used BP quantitative trait loci (QTLs) from hypertensive rats as functional proxies to seize human orthologs marked by GWAS. Results : 6 BP QTLs correspond to 6 specific human genes. BP was altered by these QTL alleles, and yet, the human non-coding GWAS variants are absent in rodents. They cannot contribute to physiological modulations of BP by these QTLs, because depleting such a variant has no impact on BP. Thus, these as a single QTL alone. Mechanistically, these QTLs may function in a common pathway. Each is involved in a different pathway step leading to BP control, not by altering BP by merely affecting QTL expressions. One pathway is muscarinic cholinergic receptor 3 (M3R) signaling. A new M3R component is implicated from current work. Conclusions: In spite of cognitive impedance from a human-centric dogma, the modularity/pathway concept is evolving into a paradigm physiologically applicable to mammalian polygenic and quantitative traits.


1.1.
A high prevalence of chronically-elevated blood pressure, hypertension, is a compelling risk for cardiovascular, renal and infectious diseases [1]. This risk has been highlighted by a recent hospitalization rate of COVID-19 patients with underlying conditions (DOI: 10.15585/mmwr.mm6915e3). The most common among them is hypertension that needs to be treated with anti-hypertensive drugs. This alarming hazard urgently demands our actions in unraveling pathogeneses of hypertension, and in distinguishing mechanistic causes of pathophysiology from their outward effects found in epidemiology.

1.2.
Thanks to genome-wide association studies (GWASs), detecting human quantitative trait loci (QTLs) for blood pressure (BP) has statistically marked the vicinity of more than 900 BP QTLs by more than 10000 single nucleotide polymorphisms (SNPs) [2]. So far, no human QTLs have been functionally identified to belong to a physiology system known to affect blood pressure. We are no closer in understanding a pathogenesis for human polygenic hypertension than before the advent of GWASs [3]. With due respect, >90% of these SNPs cannot be functional variants in spite of the equally-strong statistics associating them all with blood pressure. These SNPs are pure markers for potential QTLs close by. Thus, in identifying a human QTL, statistics are insufficient and physiological studies in vivo are needed.

A physiological distinction is unmistakable between
locating a SNP marking a QTL nearby and identifying the QTL itself. By focusing on after effects of BP-regulating mechanisms, human GWASs have reached their limitations in finding causes of these mechanisms. A QTL refers to a locus residing in a chromosome segment when genetically defined, but a QTL is a single gene when molecularly identified [4]. For example, C17QTL1 on rat Chromosome 17 is a single gene encoding Chrm3 [muscarinic cholinergic receptor 3 (M3R)] [5,6]. No combination with another QTL/gene is necessary to physiologically affect blood pressure [7,8].

1.4.
Functionally, an alteration in a physiological mechanism will cause variations in blood pressure, but not all BP variations are a result of mechanistic changes. Experimental advantages using rodent models allow causative mechanisms modulating blood pressure driven by physiology to be unveiled [9]. Because of conserved mechanisms, studying them in regulating rodent BP is equivalent to revealing the same mechanisms in humans originating from their common ancestors. Evidently, most land living mammals attain a similar range of blood pressures [10], despite differences in separate physiology characters (e.g. corporal bulk). The only way for this to materialize is that basic mechanisms modulating blood pressure must have been formed and held constant in common ancestors of these mammals before 90 million years ago (www.timetree.org), before humans existed.

In vivo studies now unify animal model and human
QTLs into a basic framework in physiological mechanisms of BP control. Rodent QTLs as proxies from inbred strains have functionally captured distinct human QTLs [11,12]. The intergenic GWAS SNP close to CHRM3 [13] is only a marker for the human QTL, not the QTL itself [11]. Thus, a 'common' SNP from human GWAS merely marks a nearby physiological QTL that has a 'rare' functional variant. Replicating such a 'common' SNP by other GWAS seems an epidemiological exercise [2] that is irrelevant to its physiological impact on BP, because removing the SNP has no effect on BP [5,7].

1.6.
These atypical and counter-intuitive results may appear confusing and disturbing, because they contradict the prevailing tenet in epidemiology that a quantitative and polygenic trait should be made by accumulating 'miniscule' effects from multiple QTLs. This is lately molded into an 'omnigenic' hypothesis [3], which has been believed to universally apply to all polygenic and quantitative traits in whatever organisms in both inbreds and oubreds. (Comparing inbreds to outbreds will be elaborated further in discussions). However, facts are facts. There is no in vivo evidence that any of human genetic architectures of GWAS in outbreds [2] could actually impact on blood pressure physiologically, and controlling gene expressions from a GWAS SNP might affect BP physiologically.

1.7.
The shift of paradigm from 'omnigenicity' to QTL modularity [11,12,14] has become a part of the literature in polygenic research [15]. Invisible QTL modularity from human epidemiology reflects limitations of GWAS [2], because GWAS is done by ignoring BP-controlling physiology and mechanisms.
Among mammals, modularity is a physiological reality [9] but hidden from GWAS. It is analogous to lacking evidence for black holes in Newton's mechanics, although they exist from Einstein's theory of gravity [9]. It seems that physiologically understanding quantitative and polygenic traits such as BP in biology mimics studying gravity in physics.

1.8.
Only by accumulating evidence can we make this paradigm shift recognizable and accepted by the concerned scientific community. In retrospect, the inherent truth embodied in Mendelism became established and appreciated after a 35-year oblivion, only when Mendel's discoveries were reproduced by others. Hopefully, our previous work [11,12,14] and the current confirmation of them, will encourage different scientists to expand one-lab-based findings in animal models and humans to broader polygenic traits including BP. In this way, the validity of this developing paradigm can be further tested.

1.9.
Arguably, a narrow range of work [11,12] might not represent a broad mechanistic reality. We have extended the rat QTL coverage to evaluate additional human GWAS genes in progressive stages, as a lone investigator-initiated lab can. Instead of replicating same QTLs in other populations as human GWASs do [3], we tested the reproducibility of mechanistic outcomes from analyzing previously-unexplored rat QTLs that respond to different human GWAS gene orthologs. In this process, our new data have validated and widened the paradigm of QTL modularity to both rodents and humans as pathogenic pathways to polygenic hypertension [9]. Previously-unsuspected components of these pathways have been implicated.

Animals
Protocols for handling as well as maintaining animals were approved by our institutional animal committee (CIPA). Inbred hypertensive Dahl salt-sensitive rats (DSS) are our functional proxy of choice. In order to detect the physiological impact from a BP QTL, our work was done in the DSS genetic background that has lost its genome buffering capacity in impeding BP fluctuations [16] and in suppressing hypertension [17,18].

Experimental protocols and analyses
Breeding procedure, dietary treatments, telemetry implantation, postoperative care and BP measurement durations were essentially the same as reported previously [14]. In brief, male rats were weaned at 21 days of age, kept on a low salt diet followed by a high salt diet starting from 35 days of age until the end of the experiment.
Telemetry probes were implanted at 56 days of age (namely 3 weeks from the time of the high salt diet). In the BP presentation (Figure 1), averaged readings of mean arterial pressures (MAP) for the duration of measurement were given for each strain.

Repeated measures' analysis of variance (ANOVA)
followed by Dunnett's test, which corrects for multiple comparisons and unequal sample sizes, was used to compare a parameter in MAP between 2 groups as reported previously (14). The power and sample size calculations in the analysis are the same as given previously [11].

Congenic knock in genetics is a proxy tool in
physiologically catching human GWAS genes by causality. The congenic principle is similar to that of SNP 'knock-in' with a variation in a genome scale [4], and is employed as congenic knock in genetics [11,12].
Despite the DSS rats are known for their 'saltsensitivity' for physiological studies, the human GWAS genes that have been captured by the DSS model commonly function in humans from general populations, with or without salt sensitivity [11,12].
Furthermore, the M3R signaling pathway found in DSS is pro-hypertensive even under low salt diet [5]. High salt diet merely accelerated hypertension and our studies of it using the model. Thus, BP-regulating mechanisms discovered in DSS rats are applicable to those of human essential hypertension in general populations, irrespective of the salt content. Among DSS chromosome segments known to contain BP QTLs [14], only those matching human QTL signals from GWAS [2] were investigated here (Supplementary Table 1).

The M3R signaling pathway in regulating BP existed in common ancestors of humans and rodents.
Since they have similar blood pressures in a polygenic and quantitative context, we hypothesized that humans and rodents may use same pathways originating from their common ancestors and their similar BP states are not due to a convergent evolution event. We tested this hypothesis by focusing on M3R as C17QTL1, a QTL on rat Chromosome 17 and human CHROMOSOME 1, because M3R has been physiologically proven to be C17QTL1 [5,7,11].

3.2.3.
This is a proof that, common ancestors of humans and rodents possessed a BP-regulating mechanism of M3R signaling pathway, in spite of the fact that M3R is pleiotropic in functions in addition to regulating BP.
This evidence explains similar BPs between humans and rodents, for which a possible convergent evolution by different mechanisms to obtain similar blood pressures has no proof. This contrast will be dealt with further in discussions. is largely due to environmental effects, not due to mechanistic actions of QTLs [9]. Since environmental factors are not inherited and mostly unquantifiable, the often-termed 'missing' heritability is not equivalent to missing total variance. Regardless how many BP QTLs really exist in an individual organism, identifying the physiological impact of a particular QTL on BP is critical that can be untangled from the nonphysiological total variance 3.3.2. The BP effect for a human GWAS SNP [2] and presumably from one QTL marked by it, seemed 'miniscule' when fractionated from total variance [3].
What then is the significance in identifying 6 'trifle' QTLs among 900 [2], each encoded by a single gene ( Chromosome 17 proves the point ( Figure 1A).

3.3.4.
Ten different human GWAS genes [2] fell into the congenic knock-in segment that defined C17QTL2 ( Figure 1A). One human GWAS signal close to CHRM3 [13] was present in the segment containing C17QTL1. The 2 QTLs were detected as a single QTL statistically and explained 6.7% of total variance in a heterogeneous rat population [20]. In contrast, each of them was capable of physiologically and independently altering BP by 28-42% in the total difference by mmHg in vivo between 2 parental rat strains and under a uniform environment ( Table 1). The calculation is as follows.

3.4.3.
In total, 49 human GWAS genes are contained in the chromosome segments harboring the 6 QTLs functional proxy, these 6 human GWAS genes may be classified physiologically into 2 epistatic modules, or 2 independent pathways in determining BP. BP is functionally additive between 2 members of 2 separate modules [14], and is the basic mechanism of QTL actions. Since M3R is a signaling pathway [5], epistatic module 2 to which C17QTL1/CHRM3 and C17QTL2 belong constitutes a pathway with multiple steps composed of different QTLs leading to BP control.
C17QTL2 most likely participates in one of these steps in the M3R signaling pathway.

3.4.4.
In order to identify a specific step in a pathway, a molecular identification of a QTL is necessary.
Identifying CHRM3 as a causal gene to C17QTL1 is the precedent for genetically discovering a component of a step in a pathway in a polygenic context [5,6]. CHRM3 has been proven to be C17QTL1 not only in DSS rats [5], but also a strong candidate for humans [11]. Chrm3 carries a function-changing missense mutation [5]. Thus, missense mutations are priority, although not exclusive, targets for identifying candidate genes for the following QTLs.
'Knocking it out' in rodents has no impact on blood pressure ( Figure 1C). Consequently, rs6141767 itself is a byproduct of primate evolution, and a human-centered marker for the functional C3QTL3 nearby, not the QTL per se.  (Table 2). These

Closely-linked
SNPs are by products of primate evolution (Table 3).
They alone or collectively do not change BP by C10QTL1 in vivo ( Figure 1C), since humans and rodents achieve similar BP with or without them [10].  Table 6), yet an intronic SNP, rs2645466, in VMP1 was found to be associated with BP. Rs2645466 is a by-product of primate evolution (Table 3), is not conserved in the rat (  Figure 1C), which is about 400kb away from VMP1. Thus, it is likely that 2 separate QTLs may exist in the C10QTL5-residing interval.

3.7.5.
In contrast to the functional correspondence of the 4 human GWAS genes to C10QTL1 and C10QTL5 ( Figure 1B), there was no BP effect by knocking in the rat ortholog of a human GWAS gene, ACE, in C10S.L8.

Thus, the relevance of ACE as a human GWAS gene in
BP regulation needs to be tested in animal models other than DSS.

QTL Modularity/pathway is the physiological framework of QTLs regulating blood pressure invented in mammalian ancestors: QTL Modularity is
the genetic framework in physiologically modulating BP embedded in their ancestral genomes [9]. The 6 QTLs and their corresponding human GWAS genes may function via only 2 modules in physiologically controlling BP and implicates 2 pathways of hypertension pathogeneses. One of them is the M3R signaling pathway [5,6]. This conservation of BPregulating pathways such as the M3R signaling supports similarity in blood pressures between differing orders of mammals [10]. As a result, the fundamental framework of BP-controlling mechanisms in pathways with multiple steps must have been established in common ancestors of mammals before they started to diverge [9].  [5]. The paradox of a 'common' SNP/marker with no effect on BP identifying a 'rare' BP-impacting variant nearby has been previously addressed in Discussion in reference [11], and will not be reiterated here.

4.1.5.
More than 10,000 human SNPs are found to be associated with over 900 genes [2].

4.2.2.
Depleting M3R diminishes vaso-relaxation that is supposed to increase BP, but contrarily decreases blood pressure [5,7]. Thus, viewing functions of these cell/tissue structures in vitro and in isolation cannot predict the actual physiological BP in vivo for a QTL.
M3R is mostly produced in the brain, less in adrenals and not detectable in heart, kidneys [5,7]. We are still puzzled [6] as to how M3R promotes hypertension, and from the brain and adrenals, modulates concordant probable blood pressure effect from a single SNP is fractionated by total variance in phenotypic variations according to Fisher [3], not its actual physiological effect on blood pressure in mmHg.

Insights into mechanistic and physiological
causes of blood pressure regulation from QTL modularity: The genetic modularity of QTLs [14,19] has broadened the scope of Mendelism to cover polygenic traits, and revealed causes and a mechanistic frame work of BP physiology [9]. Polygenicity of blood pressure is composed of individual Mendelian 'monogenic' components that are organized into modules. Mendelism is the fundamental basis for BP as a polygenic trait, and is in principle, equivalent to carbon being the basic chemical element in forming poly-carbon graphite and diamonds.

Recently, an 'omnigenic' hypothesis has been
proposed to explain GWAS results on generic quantitative phenotypes [3]. It can be described as an anthropocentric (or human-centered) theory, because non-coding GWAS SNPs only exist in humans, and not in rodents, as our previous [11] and current findings have shown. It basically proposes that regulations at gene expressions at cellular level would determine the GWAS SNPs' roles in human phenotypes including BP.
This is contrary to the modularity idea and physiological proofs underlying it [9]. to 'omnigenicity' cannot affect these pathways, because they only began to appear in primates, but do not exist in rodents (Tables 2 and 3). Conversely, rodents' noncoding Chrm3 SNPs are not conserved in humans.
Inactivating the M3R signaling pathway did not touch any of them, yet BP changed [5]. Thus, the functional M3R signaling coexists in humans and rodents, and should be shared in determining the hypertension pathogenesis, to which the non-coding rodent Chrm3 SNPs and the human-centered GWAS SNPs are irrelevant.

4.4.8.
Certain QTLs starts to function at embryogenesis [5], before the onset of adult BP physiology. Modularity can [9], whereas the 'omnigenicity' cannot, explain that a pathway involved in BP control can temporally begin at embryogenesis and continue into adulthood [5].

4.4.9.
In conclusion, modularity is supported by reproducible lines of physiological evidence as a signaling pathway, and is mechanistic in our understandings of causes for BP physiology [9]. The intuitive 'omnigenic' hypothesis [3]

Conclusions
Studies This emerging paradigm encompasses not only humans, but also most other land mammals, and is a departure from the human-centric precept [3] which is reminiscent of geocentrism distorting heliocentricy of our solar system in cosmology.

Funding
The work was supported by a grant from Heart and Stroke Foundation of Canada to AYD.