JAK2 rs10974944 is associated with both V617F-positive and negative myeloproliferative neoplasms in a Vietnamese population: A potential genetic marker.

The JAK2 gene encodes for a non-receptor tyrosine kinase that plays a key role in the JAK/STAT signaling transfer pathway. Genetic polymorphisms of this gene have been indicated to be associated with myeloproliferative neoplasm-associated thrombosis in recent studies. This research aimed to evaluate the association between the variant rs10974944 and different types of Myeloproliferative neoplasms disorders in the Vietnamese population. DNA samples were obtained from 172 essential thrombocythemia patients, 14 primary myelofibrosis patients, 76 polycythemia vera patients, and 192 healthy controls. The JAK2 rs10974944 and V617F genotypes were identified by the polymerase chain reaction-restriction fragment length polymorphism genotyping and Sanger sequencing methods. Results showed that there was a strong association between rs10974944 and Myeloproliferative neoplasms phenotype (p < .0001) and the most significant association was observed in the recessive model of the mutant allele (G). The G allele carriers had a 1.74, 2.86, and 3.03 higher risk of getting essential thrombocythemia, primary myelofibrosis, and polycythemia vera, respectively. Interestingly, this effect of rs10974944 seemed to be independent of the JAK2 V617F genotype. The distribution of rs10974944 genotypes were significantly different between V617F-positive and negative groups (p = .008). Moreover, the GG genotype of rs10974944 was observed to be associated with the risk of getting Myeloproliferative neoplasms both in JAK2 V617F-positive group, and for the first time in JAK2 V617F-negative patients. A systematic meta-analysis in different populations strengthened the evidence regarding the correlation between rs10974944 and myeloproliferative neoplasm disorders. To sum up, our results suggested that rs10974944 can be used as a predisposition screening marker for predicting Myeloproliferative neoplasms susceptibility.

BACKGROUND

Myeloproliferative neoplasms (MPNs) are a rare group of hematologic cancer, caused by an abnormality of the mature peripheral blood cell production in the bone marrow. These diseases were reported to cause a higher mortality rate compared to healthy controls: In 5 years, the survival rate of MPNs patients was 55% compared to 90% in the matched control (Hultcrantz et al., 2015). MPNs are categorized into two groups: the Philadelphia chromosome‐positive (Ph‐positive), which is associated with the translocation t(9;22) (q34;q11) of the BRC (OMIM: 151410) gene in chromosome 22 and the ABL1 (OMIM: 189980) gene in chromosome 9 to form the fusion gene called BCR‐ABL1; and the negative Philadelphia group (Ph‐negative), which includes three disorders: essential thrombocythemia (ET), polycythemia vera (PV), and primary myelofibrosis (PMF) (Arber et al., 2016).

ET is characterized by a high platelet production in peripheral blood by megakaryocytes (450 × 109/L or higher). The prevalence of this disorder among US citizens was about 38 to 57 patients in 100,000 people (Tefferi, 2014). PV is the uncommon condition of bone marrow that overproduces erythrocytes, with hemoglobin levels greater than 16.5 g/dL in males and 16.0 g/dL in females. In the US, prevalence of PV was estimated to be approximately 44–57 per 100,000 of the population (Mehta et al., 2014). In PMF, the abnormal proliferation of hematopoietic stem cells in bone marrow causes fibrosis and scar tissue and thus, bone marrow cannot produce enough normal blood cells (Tefferi, 2014). Many studies have indicated that the host genetic factor might play a critical role in the risk of getting MPNs. Among those, the abnormal of the JAK/STAT signaling interaction pathway are considered as the most important criterion of MPNs pathogenesis with many driving mutations belonging to different genes such as JAK2 (OMIM: 147796), MPL (OMIM: 159530), CALR (OMIM: 109091) and many other genes (Greenfield et al., 2021; Viny & Levine, 2014). Especially, the missense mutation JAK2 V617F or NM_004972.4 (JAK2):c.1691G > T p.Arg564Leu (rs77375493) was identified as the most common mutation leading to MPNs as this mutation was detected in 95% in patients with PV, 50% to 70% in ET, and 40% to 50% in PMF (Vainchenker & Kralovics, 2017).

In this study, we investigated the association of the variant JAK2 rs10974944 with 262 MPNs patients and 192 healthy controls in a Vietnamese population. We also compared the allele frequency of rs10974944 in each JAK2 V617F genotype to investigate the linkage disequilibrium between these two variants. Finally, we performed a meta‐analysis to understand more clearly the correlation between rs10974944 and the risk of MPNs in different populations. These initial data can be used for further studies on exploring the progression of MPNs and their application in the genetic diagnosis of the disease.

MATERIALS AND METHODS

Studied population

The studied population included 262 MPNs patients (172 of ET, 14 of PMF, and 76 of PV patients), and 192 healthy controls, recruited at the Vietnam Military Medical University during 2018–2019. All participants were well informed about the research and signed the informed consent. This research was approved by the Institutional Review Board of the Institute of Genome Research, Vietnam Academy of Science and Technology (No 4‐2021/NCHG‐HDDD).

Variant genotyping

Total peripheral blood samples of all the participants were collected for gDNA extraction by the QIAamp® DNA Mini Blood Kits (QIAGEN) according to the manufacturer's protocol. The accession number of JAK2 Genbank reference sequence is NM_004972.4. Genotypes of the two genetic markers JAK2 V617F (rs77375493) and rs10974944 were identified by the polymerase chain reaction‐restriction fragment length polymorphism (PCR‐RFLP) method using specific primers (Table 1). After size verification by 1% agarose gel, the PCR products were digested with the BsaXI and BclI restriction enzymes for the variant rs77375493 and rs10974944, respectively. The genotype of each individual was accordingly determined by electrophoresis on 3% agarose gel. And 10% of the obtained genotypes were sent to sequenced by the Sanger method at Apical Scientific Laboratory (Malaysia) to verify the result was identical. The genotypes of the two markers were screened at the same time.

TABLE 1

Primers and restriction enzymes (RE) used for PCR‐RFLP

Name of oligo	Sequence	PCR product size	RE
rs77375493‐FP	TCCTCAGAACGTTGATGGCAG	453 (bp)	BsaXI
rs77375493‐RP	ATTGCTTTCCTTTTTCACAAGAT
rs10974944‐FP	ACATGGGTTTTGCATCCTATGAA	492 (bp)	BclI
rs10974944‐RP	TCTGCTTGCTAGTGGGTGAAT

Linkage disequilibrium and statistical analysis

The linkage disequilibrium and statistical analysis were implemented using the R software version 4.0.2 and Rstudio. Linkage disequilibrium was assessed between rs77375493 and rs10974944 using the LDlinkR package of R language (Myers et al., 2020). The associations between genotypes or allele groups and MPNs phenotypes were checked using the Chi‐squared test and Fisher's exact test, accordingly. The odds ratios (OR) and 95% confidence intervals (95% CI) were calculated using Microsoft Excel following the formula by Szumilas et al (Szumilas, 2010). All of the statistical tests were applied as two‐sided. An obtained p‐value less than 0.05 was considered statistically significant.

Meta‐analysis

The literature search was conducted on the association between rs10974944 and MPNs phenotype using Pubmed database (https://pubmed.ncbi.nlm.nih.gov/). All the Review papers, Case reports, replication studies on the same population, researches on un‐relevant diseases were excluded from further analysis. The meta‐analysis was carried out by METAL (Willer et al., 2010). Potential publication bias in meta‐analyses was identified by the Egger's test.

RESULTS

Population characteristics

All 454 participants involved in this study were Vietnamese people belonging to the Kinh ethnicity group. The gender ratio (Male/Female) of the studied population was 0.76 (196 Males and 258 Females) and the age mean of all the population was 56.32 ± 11.38. Detailed information on the gender and age structure of each group was listed in Table 2. There was no observed significant difference between the MPNs group and the healthy group in term of Age, Gender, or Ethnicity.

TABLE 2

Characteristics between MPNs patients and healthy controls

Group	Number N	Age mean ± St.d	Gender male/female (% of male)	Ethnic
Control	192	52.61 ± 9.52	88/104 (45.8%)	Kinh (100%)
ET	172	53.65 ± 12.58	60/112 (34.9%)	Kinh (100%)
PMF	14	61.64 ± 8.75	4/10 (28.6%)	Kinh (100%)
PV	76	58.16 ± 12.11	44/32 (57.9%)	Kinh (100%)
p‐value		0.39 1	0.327 2	N/A

p‐value obtained by Mann–Whitney U test.

p‐value obtained by Chi‐squared test.

The distribution of JAK2 rs77375493 and rs10974944

In this studied population, 161 individuals carried the JAK2 V617F mutation (account for 35.5% of all population and 61.5% of MPNs patients). Among those, 99/172 (57.6%) of the ET patients, 9/14 (64.3%) of the PMF patients, 53/76 (69.7%) of the PV patients and none of the Control group were JAK2 V617F positive.

The result of rs10974944 genotyping showed that the genotype distribution (CC/GC/GG) in all the population was 177/173/104, in Control group was 90/82/20, 63/64/45 in ET patients, 4/4/6 in PMF patients and 20/23/33 in PV patients. The correlation between JAK2 rs10974944 and JAK2 V617F mutation status was shown in Table 3. The linkage disequilibrium analysis indicated that there was a slight linkage disequilibrium between these two variants, though it was not tight linkage (D′ < 0.8) (Figure 1). If only the MPNs patients were taken into account (excluded the controls), there was a significant difference in the genotype distribution of rs10974944 between the V617F‐positive and negative group: the frequency of GG genotype was higher in the V617F‐positive group (38.5% compared to 21.8%), the frequency of CC was lower (27.3% compared to 42.6%) while the heterozygous genotype GC was the same (34.2% and 35.6%) (p = .008).

TABLE 3

Genotype distribution between the two variants JAK2 V617F (rs77375493) and rs10974944 in each group

rs10974944 genotype	JAK2 V617F negative			JAK2 V617F positive
	CC	GC	GG	CC	GC	CC
Control	90	82	20	0	0	0
Essential thrombocythemia	31	27	15	32	37	30
Primary myelofibrosis	2	2	1	2	2	5
Polycythemia vera	10	7	6	10	16	27
All population	133	118	42	44	55	62

The difference in the genotype distribution of rs10974944 between the V617F‐positive and negative group was significant with p = 0.008 are indicated in bold.

FIGURE 1

Linkage study consisted of JAK2 rs77375493 and rs10974944. D′ value was shown in the LD block

Association of rs10974944 with MPNs in the studied population

Statistical analysis showed that the genotype of rs10974944 was associated with ET, especially in the additive model (p = .000421) and in the recessive model (OR = 3.047, 95% CI = 1.716–5.413, p = .00009). The G allele of this variant also increased the risk of getting ET in comparison to the C allele (OR = 1.74, 95% CI = 1.286–2.354, p = .0003) (Table 4). Similarly, this polymorphism showed a significant link with PMF and PV risk. The most significant model was also the recessive model (OR = 6.45, 95% CI = 2.03–20.48, p = .0004 in PMF; OR = 6.6, 95% CI = 3.452–12.62, p ~ 10−9 in PV). The frequencies of allele G were significantly higher in the patient groups compared to control (OR = 2.86, 95% CI = 1.31–6.24, p = .006 in PMF; OR = 3.03, 95% CI = 2.06–4.47, p ~ 10−8 in PV) (Tables 5 and 6).

TABLE 4

Association of rs10974944 with essential thrombocythemia

	Control (n = 192)	Case (n = 172)	Odd ratio	95%CI	p‐value
Addtive model
CC	90 (46.9%)	63 (36.6%)	1.00		0.000421
GC	82 (42.7%)	64 (37.2%)	1.115	0.705–1.764
GG	20 (10.4%)	45 (26.2%)	3.214	1.734–5.959
Dominant model
CC	90 (46.9%)	63 (36.6%)	1.00		0.048
GC + GG	102 (53.1%)	109 (63.4%)	1.527	1.003–2.324
Recessive model
CC + GC	172 (89.6%)	127 (73.8%)	1.00		0.00009
GG	20 (10.4%)	45 (26.2%)	3.047	1.716–5.413
Overdominant model
CC + GG	110 (57.3%)	108 (62.8%)	1.00		0.285
GC	82 (42.7%)	64 (37.2%)	0.795	0.522–1.21
Alleles
C	262 (68.2%)	190 (55.2%)	1.00		0.000308
G	122 (31.8%)	154 (44.8%)	1.74	1.286–2.354

TABLE 5

Association of rs10974944 with primary myelofibrosis

	Control (n = 192)	Case (n = 14)	Odd ratio	95%CI	p‐value
Addtive model
CC	90 (46.9%)	4 (28.6%)	1.00		0.00776
GC	82 (42.7%)	4 (28.6%)	1.098	0.266–4.533
GG	20 (10.4%)	6 (42.9%)	6.75	1.741–26.16
Dominant model
CC	90 (46.9%)	4 (28.6%)	1.00		0.267
GC + GG	102 (53.1%)	10 (71.4%)	2.206	0.669–7.278
Recessive model
CC + GC	172 (89.6%)	8 (57.1%)	1.00		0.000418
GG	20 (10.4%)	6 (42.9%)	6.45	2.031–20.48
Overdominant model
CC + GG	110 (57.3%)	10 (71.4%)	1.00		0.3
GC	82 (42.7%)	4 (28.6%)	1.864	0.565–6.152
Alleles
C	262 (68.2%)	12 (42.9%)	1.00		0.006029
G	122 (31.8%)	16 (57.1%)	2.863	1.314–6.237

TABLE 6

Association of rs10974944 with polycythemia vera

	Control (n = 192)	Case (n = 76)	Odd ratio	95%CI	p‐value
Addtive model
CC	90 (46.9%)	20 (26.3%)	1.00		6.35 × 10−9
GC	82 (42.7%)	23 (30.3%)	1.262	0.646–2.466
GG	20 (10.4%)	33 (43.4%)	7.425	3.553–15.52
Dominant model
CC	90 (46.9%)	20 (26.3%)	1.00		0.002043
GC + GG	102 (53.1%)	56 (73.7%)	2.471	1.378–4.430
Recessive model
CC + GC	172 (89.6%)	43 (56.6%)	1.00		9.7 × 10−10
GG	20 (10.4%)	33 (43.4%)	6.6	3.452–12.62
Overdominant model
CC + GG	110 (57.3%)	53 (69.7%)	1.00		0.06
GC	82 (42.7%)	23 (30.3%)	1.718	0.975–3.028
Alleles
C	262 (68.2%)	63 (41.4%)	1.00		1.06 × 10−8
G	122 (31.8%)	89 (58.6%)	3.034	2.059–4.471

Result of statistical analysis also indicated that rs10974944 associated with both JAK2 V617F‐negative MPNs and V617F‐positive MPNs in this studied population. In the V617F‐negative MPNs patients group, frequency of the GG genotype was twice as high compared to control (21.8% to 10.4%) (OR = 2.395, 95% CI = 1.24–4.64, p = .0097). Among the V617F‐positive MPNs patients, frequency of GG genotype was even four times higher than control (38.5% to 10.4%) (OR = 5.3859, 95% CI = 3.07–9.44, p < .0001). In term of allele effect, while the G allele of rs10974944 seemed to dramatically increase the risk of getting MPNs in the V617F‐positive group (OR = 2.69, 95% CI = 1.98–3.66, p < .0001), this effect was no longer statistically significant in the V617F‐negative group (OR = 1.41, 95% CI = 0.99–2.01, p = .0584).

Meta‐analysis on the association of rs10974944 with MPNs

After all the unsuitable articles were excluded from the literature mining, a total of 7 studies were eligible for the final pooled analysis. The related information of 7 included studies and data from this study were represented in Table 7. The sample size of each study was from 146 to 962 participants, with the average size as 486 individuals. Three studies were conducted in the European population, three studies in the Asian population, 1 in the South American population and 1 study with an unknown population.

TABLE 7

The general characteristics of studies included in the meta‐analysis

Author	Population	Year	MPN patients			Control			H‐W p‐value	Ref
			GG	GC	CC	GG	GC	CC
Pagliarini‐e‐Silva et al	Brazilian	2013	18	20	18	12	25	53	0.0046	Pagliarini‐e‐Silva et al. (2013)
Koh et al	Chinese	2014	29	76	23	40	198	232	0.8061	Koh et al. (2014)
Matsuguma et al	Japanese	2019	55	82	64	33	127	206	0.0419	Matsuguma et al. (2019)
Hsiao et al	unknown	2011	10	26	25	8	46	52	0.6168	Hsiao et al. (2011)
Zerjavic et al	Slovenian	2013	12	50	73	41	164	254	0.0558	Zerjavic et al. (2013)
Soler et al	Spanish	2015	18	62	49	19	115	136	0.4227	Soler et al. (2015)
Trifa et al	Romanian	2016	88	281	160	35	171	227	0.7258	Trifa et al. (2016)
This study	Vietnamese	2021	84	91	87	20	82	90	0.8365

In general, the meta‐analysis showed a statistically significant relation between the JAK2 rs10974944 genotype and the risk of getting MPNs disorders (OR = 1.908, 95% CI = 1.529–2.381, p = 10−8) under the random‐effects model (Figure 2). Although 6/8 studies obtained significant results, the analysis detected a potential evidence of heterogeneity in the results for the above association (I 2 = 77%, τ2 = 0.075, p‐value for heterogeneity <.01).

FIGURE 2

Forest plot demonstrated the association between JAK2 rs10974944 and MPNs susceptibility

In term of publication bias, a review of the funnel plot indicated no potential for publication bias as no signs of asymmetry or hole was observed (Figure 3). The Egger's test confirmed that there was no clear evidence of publication bias (p‐value for Egger = .709).

FIGURE 3

Funnel plot of publication biases on the association between JAK2 rs10974944 and MPNs

DISCUSSION AND CONCLUSION

The JAK2 gene encodes the Janus kinases 2, one of the four members of the JAK‐family (JAK1, JAK2, JAK3, and TYK2) – the non‐receptor tyrosine kinase that activates cytokine‐mediated signals by the JAK–STAT pathway (Sopjani et al., 2021). JAK2 rs10974944 is part of the JAK2 46/1 haplotype, which consists of four main SNPs (rs10974944, rs1159782, rs3780367, and rs12343867) and hundreds of other SNPs located in the three genes: Insulin‐like 4 (INSL4), Insuline‐like6 (INSL6), and JAK2 (Olcaydu, Harutyunyan, et al., 2009; Olcaydu, Skoda, et al., 2009). These four SNPs were indicated in complete linkage disequilibrium with each other in many populations (Anelli et al., 2018; Tanaka et al., 2013), thus they were also referred to as “GGCC” haplotype as their mutant alleles.

In this study, results showed that the rs10974944 was strongly associated with MPNs in the Vietnamese population. In detail, the G allele of rs10974944 significantly increased the chance of getting ET, PMF, and PV disorder 1.74, 2.86, and 3.03 times, respectively. This result seemed to be consistent with other publications on the association between JAK2 rs10974944 and MPNs phenotypes worldwide as our meta‐analysis data showed a significantly different distribution of rs10974944 genotypes between MPNs patients and healthy controls (OR = 1.908, 95% CI = 1.529–2.381, p = 10−8).

Several previous studies found that rs10974944 was associated with the development of V617F‐positive MPNs (Jones et al., 2009; Kilpivaara et al., 2009; Tanaka et al., 2013). This observation might be due to the linkage between the JAK2 46/1 and V617F as many studies had demonstrated that this haplotype was associated with V617F‐positive myeloproliferative neoplasms in Brazilian (Macedo et al., 2015), Romanian (Trifa et al., 2010), Japanese (Tanaka et al., 2013) and other populations (Anelli et al., 2018; Stolyar et al., 2018). However, the results of our study indicated that in the Vietnamese population, the linkage disequilibrium between rs10974944 and rs77375493 (V617F) was not tight, yet rs10974944 still had a strong association with MPNs, suggested that the effect of rs10974944 on MPNs phenotype was independent of V617F profile. In addition, our research observed a significant association between the rs10974944 GG genotype and the risk of getting V617F‐negative MPNs. Up to our knowledge, this correlation is novel and has not been published in any other population.

In conclusion, in this study, we detected an association between JAK2 rs10974944 and MPNs in a Vietnamese population. This association seemed to be independent with the JAK2 V617F genetic profile of the MPNs patients, since the correlation between rs10974944 and MPNs was observed in both V617F‐positive and V617F‐negative groups compared to controls. Our result was confirmed by a meta‐analysis of 7 other studies on Brazilian, Chinese, Japanese, Slovenian, Spanish, Romanian populations. Further functional analysis should be implemented to investigate the possible pathogenic role of rs10974944 as well as the haplotype JAK2 46/1 on the MPNs phenotypes.

AUTHOR CONTRIBUTIONS

Nguyen Thy Ngoc (N.T.N) designed the study. N.T.N received the grant for the study. Nguyen Ba Vuong (N.B.V) and Nguyen Thi Xuan (N.T.X) carried out the sampling. Bui Bich Hau (B.B.H) and N.T.N carried out the laboratory work. N.T.N analyzed the data and wrote the manuscript, with input from all authors.

CONFLICT OF INTEREST

No potential conflict of interest was reported in this study.

ETHICS STATEMENT

The authors stated that they have obtained the appropriate institutional review board approval and followed the principles outlined in the Declaration of Helsinski for all human or animal experimental investigations. In addition, for investigations involving human subjects, informed consent has been obtained from the participants involved.