|Home | About | Journals | Submit | Contact Us | Français|
Using congenic strains of the Dahl salt-sensitive S rat introgressed with genomic segments from the normotensive Lewis rat, a blood pressure quantitative trait locus was mapped within 104kilobases on chromosome 10. The goal of the current study was to conduct extensive phenotypic studies and further fine-map this locus. At 14 weeks of age, the blood pressure of the congenic rats fed a low salt diet, was significantly higher by 47mm Hg (p<0.001) compared to that of the S. A time-course study showed that the blood pressure effect was significant from very young ages of 50–52 days (13mm Hg, p<0.01). The congenic strain implanted with electrocardiography transmitters demonstrated shorter QT-intervals and increased heart rate compared with S (p<0.01). The average survival of the congenic strain was shorter (134 days) compared with S (175 days, p<0.0007). The critical region was narrowed to <42.5kilobases containing 171 variants and a single gene, Rififylin. Both the messengerRNA and protein levels of rififylin were significantly higher in the hearts of the congenic strain. Overexpression of rififylin is known to delay endocytic recycling. Endocytic recycling of fluorescently labeled holotransferrin from cardiomyocytes of the congenic strain was slower than that of S (p< 0.01). Frequency of cardiomyocyte beats in the congenic strain (62±9 beats/min) was significantly higher than that of the S (24 ± 6 beats/min, p<0.001). Taken together, our study provides evidence to suggest that early perturbations in endocytic recycling caused by the overexpression of Rffl is a novel physiological mechanism potentially underlying the development of hypertension.
Previously, a blood pressure (BP) quantitative trait locus (QTL) was located within 104kb on rat chromosome 10 using congenic strains developed by introgressing alleles from the normotensive LEW (Lewis) rat onto the genome of the hypertensive salt-sensitive (S) rat . The BP QTL alleles within the 104kb region from the normotensive LEW rat are pro-hypertensive because they further increase the BP of the hypertensive S rat . The purpose of the current study was to (1) further narrow the BP QTL region, (2) expand the phenotyping and (3) study the function of the candidate gene/s.
The importance of the study is that the QTL was further fine-mapped within a shorter segment of <42.5kb containing a single gene, Rififylin or RING Finger and FYVE-like domain-containing protein (Rffl). By overexpressing Rififylin in HeLa cells, Coumailleau  et al have demonstrated that Rififylin delays recycling of membrane components from the endocytic recycling compartment to the plasma membrane. Rififylin is also an endosome-associated ubiquitin ligase 3 . The current study presents data to suggest that Rififylin is a novel gene linked to increased heart rate, BP and mortality in the congenic strain compared with S.
All animal experiments were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and as per approved protocols by the Institutional animal care and use review committee of the University of Toledo College of Medicine. The primary congenic strain described in this study is S.LEW(10)×12×2×3×5.
Rats were weaned at 30 days of age and administered low salt diet (0.3% NaCl, Harlan Teklad) for 30 days. Systolic blood pressure (BP) was measured using the tail-cuff method as described previously. The day after the last BP measurements, rats were euthanized and whole heart weights were recorded. Blood pressure was also confirmed by two independent radiotelemetry experiments using the equipment, probes and software from Data Sciences International (DSI), as described previously . Experiment 1) S (n=10) and congenic (n=10) rats on a low salt diet at the age of 75–78 days were surgically implanted with C40 blood pressure transmitters; Experiment 2) C10 blood pressure transmitters were implanted into 30 day old rats.
All phenotypic data obtained from the two groups (congenic and S rats) were statistically analyzed by Student’s t-test except the time course analyses of blood pressure by telemetry, which was assessed by two way ANOVA followed by Tukey’s post hoc test. A p-value of <0.05 was considered statistically significant. Statistical analyses of the microarray data were performed with RMA (Robust Multiarray Averaging) and BH (Benjamini and Hochberg) adjustment using the R statistical package (version 2.8.1).
Tail DNA from adult rats and mRNA from hearts of neonates and 53 days old rats were used for genomic and cDNA analyses, respectively, as described under the expanded online methods section (Please see http://hyper.ahajournals.org). Transcript expression of Rffl was analyzed by Real Time PCR (BioRad) and expression levels relative to Gapdh were calculated by the 2−ΔΔCT method .
Western blotting experiments and the sources of antibodies are detailed in the online methods section (Please see http://hyper.ahajournals.org).
RNA was isolated from the hearts of concomitantly raised, male, 53 day old S and congenic rats (n=6/group) using TRIzol and purified by RNeasy kit (Qiagen). RNA from two animals was pooled. Three such RNA samples from S and congenic rats were hybridized to Affymetrix Rat Expression Arrays 230 2.0. Statistical analyses of the microarray data were performed using R statistical package (version 2.8.1). The microarray data are in compliance with the Minimum Information About Microarray Experiments (MIAME) and are available at the Gene Expression Omnibus (GEO) database (GSE23643).
Transferrin recycling was studied as described previously  and details are provided in the online methods section (Please see http://hyper.ahajournals.org). Mean fluorescent intensity was measured in Image J at individual time points of the acquired images.
Cardiomyocytes from S and congenic rats were grown on glass cover slips for 4–7 days and placed on a stage-mounted microscope. Cardiomyocytes were chosen at random and cell beats were visually counted  and data analyzed by a two-tailed, unpaired t-test.
The electrocardiogram (ECG) was monitored on a continuous basis with a biotelemetry system (DSI). The transmitter devices were surgically implanted in 32 day old rats under general anesthesia. The devices were inserted into the peritoneal cavity and electrodes from the transmitter were arranged in Lead II configuration. ECG data were collected at 5 minute intervals for 24 hours and analyzed using Ponemah v.4.90 software (DSI). Normalization of QT interval was conducted with the Bazett’s formula specifically for rats. T-test (available within the Ponemah software) was used for statistical analysis of the ECG data.
To fine-map the BP increasing locus the congenic strain S.LEW×12×2×3×5, this strain was backcrossed to S. F1 rats were intercrossed to obtain 1995 F2 rats. None of these F2 animals were recombinants.
Using newly identified single nucleotide polymorphisms between S and LEW (Fig S1, Table S1, please see http://hyper.ahajournals.org) the previously located 104kb QTL region was resolved to <42.5kb between the base pairs 71028112 and 71070581. Sequencing of this <42.5kb interval from S and S.LEW×12×2×3×5 resulted in the identification of 171 variants (Table S2, please see http://hyper.ahajournals.org). The critical <42.5kb region contained a single predicted gene, Rififylin (Fig 1). The homologous region of this segment in humans also contains Rififylin as the single gene (Fig 1, www.ncbi.nlm.nih.gov).
The BP increasing effect of LEW alleles of the BP QTL was previously reported after feeding high salt (2% NaCl) to the rats . To test whether the hypertension increasing property conferred by this highly resolved genomic segment is independent of dietary salt S and S.LEW×12×2×3×5 congenic rats were fed with a low salt (0.3% NaCl) diet. BP measured by tail-cuff of the congenic strain (225 ± 3 mm Hg, n=10) was higher by 46 mm Hg than that of the S (178 ± 4 mm Hg, p<0.0001) (Fig 2a). Relative heart weight of the congenic strain (4.69± 0.10) was also significantly higher than that of the S (3.91±0.11) (p<0.0001). At 93 and 110 days, the hearts of both congenic and S rats were assessed by echocardiography. The detailed results of the comparisons are presented in Table 1. Representative two-dimensional and M mode images in Figure 2b showed that the congenic strain exhibited concentric cardiac hypertrophy as evidenced by increased wall thickness and reduced LV volume. At this time point, cardiac output of the congenic rats was also higher than that of the S (Table 1).
A time course study with radiotelemetry corroborated this result and demonstrated that both systolic and diastolic BP were higher in the congenic rats (Fig 2c, Table S3, please see http://hyper.ahajournals.org).
In order to determine whether the changes in BP precede the alterations in cardiac function, both systolic and diastolic BP of young rats (43 days) were monitored by radiotelemetry (Fig 2d, Table S4, please see http://hyper.ahajournals.org). At young ages of 50–52 days of age, both systolic and diastolic BP of the congenic rats were significantly higher than that of the S (P<0.01, Figure 2d, Table S4, please see http://hyper.ahajournals.org). However, there was no change in cardiac function at this time point of 50–52 days (Table S5, please see http://hyper.ahajournals.org). The difference in systolic BP between S and the congenic strain was further amplified with time, reaching a significant difference of >25 mm of Hg when the rats were 109–110 days old (Fig 2d). The increased BP of the congenic strain was associated with significantly shorter survival compared with the hypertensive S (Fig 2e). The median survival of the congenic strain was 134 days whereas that of the S was 175 days (p<0.0007).
The interval between the start of the QRS complex and end of the T wave called as the QT-interval of a heart beat was recently reported in a genome-wide study to be associated with single nucleotide polymorphisms around the human RFFL gene . This association was also confirmed in a second genome-wide association study [10, 11]. To assess whether the functional alterations caused by Rffl in our model are similarly associated with QT-interval, we surgically implanted electrocardiogram transmitters into 45 day old S and congenic rats and monitored their ECG. As shown in Figures 3a and 3b, the QT interval of the congenic strain was significantly shorter than that of the S. While the QT interval which is a useful index of the ventricular action potential duration, the action potential duration varies directly with heart rate (an increase in heart rate shortens the action potential and vice versa). Thus, QT interval also varies with heart rate and could be corrected for this variable using the Bazett’s correction. QTcb, which applies the Bazett’s correction for QT interval, was also significantly decreased in the congenic strain compared with S (Fig 3c).
To assess the candidacy of the single predicted gene, Rififylin, we sequenced all exons of Rffl, five of which were located within the <42.5kb region. There were no exonic variants. We then explored for differential transcription of Rffl as the next possibility. A microarray experiment with small sample numbers of n=3 rats/group suggested a trend for increased expression of Rffl in the hearts of congenic rats compared with S (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=zjonleysiyeqing&acc=GSE23643). cDNA products of Rffl obtained by RT-PCR (Fig 4a) were further confirmed by Realtime PCR (n=6 rats/group) (Fig 4b). The mRNA expression of Rffl was significantly higher in the congenic strain compared with S (Fig 4b). Accordingly, Rffl protein was also higher in the hearts of congenic strain compared with S (Fig 4c).
As shown in Figure 4d, the total polyubiquitinated proteins from the hearts of the congenic strain were significantly higher than that of the S (p<0.01). Microarray data of these hearts showed that only six transcripts were significantly differentially expressed (Table 2). Among these, two transcripts were prominently downregulated in the congenic strain. These were methyl CpG binding protein 2 (Mbd2), which is a suppressor of transcription  and c18orf54, a transcript of unknown function. Because the ubiquitin-proteasome system influences transcription[13–15], we reasoned that the downregulation of Mbd2 in the heart is perhaps a consequence of the increased polyubiquitination of cardiac proteins in the congenic strain.
Rififylin was overexpressed in the hearts of neonatal congenic rats (Fig 5a). Overexpression of Rififylin is described to cause delayed endocytic recycling . Therefore we quantitated endocytic recycling in isolated cardiomyocytes by the transferrin recycling assay. Recycling was significantly slower in the congenic compared with S (Fig 5b, c). After 30 minutes of incubation, % fluorescence retained in cardiomyocytes from congenic rats (29%) was higher than that in the S (12%, p <0.01) (Fig 5b and c). The frequency of cardiomyocyte beats from the congenic strain (62 ± 9 beats/min) was significantly faster than that of the S (24 ± 6 beats/min, p<0.001) (Fig 5d). Heart rate of the congenic rat (432± 17) was significantly higher compared with S (399± 10) (p<0.05).
The data presented in our study suggest that a BP QTL is further fine-mapped within <42.5kb containing a single candidate protein coding gene, Rififylin and 171 candidate quantitative trait nucleotides. The congenic strain containing LEW alleles at this QTL region has elevated BP, develops cardiac hypertrophy and has a shorter QT interval. These phenotypes are associated with an elevated mRNA and protein expression of rififylin for unknown reasons. There are two known functions of overexpression of both the transcript and protein product of the gene, Rififylin: 1) delayed recycling of endosomes  and 2) increased polyubiquitination of proteins . Both of these functions were differential between the congenic strain and S. The delay in recycling of endosomes was demonstrated in neonatal cardiomyocytes of the congenic strain. These physiological differences were associated with alterations in beats/min of neonatal cardiomyocytes from the congenic animals which also had shorter QT intervals and increased heart rates. These early changes in cardiac function are known predictors of hypertension [16, 17] and observed to occur before significant differences in blood pressure are detected between the congenic strain and S. Taken together, the data suggests that overexpression of Rffl represents one of the mechanistic factors likely to be contributing to the observed increase in blood pressure of the congenic strain. Cardiac hypertrophy observed later in the study is likely influenced by the higher blood pressure, but the concentric nature of the hypertrophy remains intriguing.
The contractile function of cardiomyocytes is primarily dependent on the surface expression of ion channels . While mutations in many ion channels per se are discovered to cause alterations in myocardial repolarization [19–21], it is possible that functional alterations in proteins involved in the intracellular transport of the channels to and from the plasma membrane could also affect membrane repolarization by regulating the dynamic control of their endocytosis and exocytosis from intracellular storage compartments. One such storage compartment is the endocytic recycling compartment (ERC), a peri-nuclear collection of tubular organelles that mature from sorting endosomes . An independent study has demonstrated that the recycling of endosomes is delayed in the presence of an excess of rififylin . We have demonstrated that this function of rififylin is recapitulated in cardiomyocytes from the congenic strain which overexpress rififylin. This observation lead us to suspect alterations in myocardial repolarization, which was further supported by the shorter QT-intervals observed in the congenic strain because QT-intervals reflect myocardial repolarization as measured by the electrocardiogram.
Rififylin is also reported to be an endosome-associated ubiquitin ligase 3 . Therefore we assessed the polyubiquitination status of cardiac proteins. Higher levels of polyubiquitinated proteins observed in the congenic strain compared with the S is suggestive of increased availability of endosome-associated ubiquitin ligase 3 activity as a result of increased expression of Rffl. Because the ubiquitin-proteasome system influences transcription [13–15], we reasoned that the observed downregulation of Mbd2 is perhaps a consequence of the increased polyubiquitination of cardiac proteins in the congenic strain compared with S.
An obvious question that arises from our current study is, ‘why is Rffl overexpressed in the congenic strain compared with S?’ The various possibilities include the following:(1) exonic or intronic variants within the Rffl gene, (2) variants within the promoter of Rffl, (3) variants upstream and downstream of Rffl that are binding sites for transcription enhancers/suppressors, (4) variants in non-coding RNAs, (5) variants within yet undefined regulatory elements of Rffl and (6) Trans-elements on the remainder of the S rat genome that interact with LEW alleles within the <42.5kb QTL region to regulate the expression of Rffl. Sequencing has revealed that while there are no exonic variants of Rffl, there are 171 candidate variants between S and LEW within the <42.5kb region. To begin addressing the question of why Rffl is overexpressed in the congenic strain, a detailed follow-up study of each of these candidate variants is required, but this is beyond the scope of the current study which was primarily focused on extended phenotyping. Nevertheless, the importance of the current study is that it provides evidence for aberrant endocytic recycling in cardiomyocytes to be further considered as a novel physiological mechanism potentially underlying blood pressure regulation.
Several mapping studies point to candidate genes on human chromosome 17 as plausible for blood pressure control in humans . Two independent genome-wide association studies (GWAS) [10, 11] also point to single nucleotide polymorphisms (SNPs) on human chromosome 17 that are associated with systolic BP. Further, one of these groups conducted a large meta-analysis of three GWAS in 13,685 individuals of European ancestry from the Framingham Heart Study, the Rotterdam Study and the Cardiovascular Health Study, as part of the QTGEN consortium and identified association of multiple minor alleles near the RFFL locus in humans with shorter QT-intervals . This association was also confirmed in a second study . Our finding of a genomic segment containing rififylin similarly linked to QT-intervals in rats serves as a functional validation of these GWAS [10, 11]. Further, our observation in rats that early changes in QT-interval contributed to the development of hypertension suggests that these individuals could be at risk for developing hypertension. Interestingly, over 29% of the cohorts screened by GWAS  were hypertensive, thus lending support to our interpretation. Overall, our studies provide the impetus to target the endosomal recycling machinery via Rffl as a novel mechanism for prevention of aberrant heart function leading to hypertension in individuals with abnormal heart rhythms.
This work was conducted through funds provided by the NIH to BJ (HL020176, HL076709). EEM acknowledges a postdoctoral fellowship award from the American Heart Association.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.