Implication of HLA-C and KIR Alleles in Human Papillomavirus Infection and Associated Cervical Lesions

Abstract

Human papillomavirus (HPV) regulation of host immune response leads to cervical lesions. In particular, natural killer (NK) cells are crucial for HPV control. Since specific HLA-I/KIR interactions modify NK cell activation, we analyzed HLA-C and KIR alleles in HPV infection and lesion development in 150 controls, 33 condyloma acuminatum, and 111 invasive cervical cancer (ICC) patients. We showed an increase in HLA-C1/KIR2DL2 and HLA-C1/KIR2DL3 pairs in HPV high-risk infected patients (OR 3.05, 3.24) with ICC (OR 1.33, 3.68). These data suggest HLA-C and KIR typing as risk marker for HPV infection and lesion evolution.

High-risk human papillomavirus (HPV) infection is the most significant risk factor for the development of premalignant and malignant epithelial lesions of the cervix (12). The HPV types that infect the genital mucosa can be divided into two groups: high-risk HPV types 16, 18, 31, and 45, frequently found in cervical cancers; and low-risk HPV types 6 and 11 that infect the genital epithelia but are rarely detected in malignancies. The development of cervical cancer starts with viral infection but requires other factors for cell transformation, as an impaired host immune system. However, the exact causes that lead to tumor have not been fully understood. Recent but few studies suggest a relationship between HLA-C alleles and HPV-related cervical diseases (2,6,5,10,11). Since we previously documented an important implication of natural killer (NK) cell inhibitory receptors (KIRs) and their interaction with HLA ligands in viral infection control in a cohort of multiple sclerosis patients (7), we focused our attention on HLA-C and KIR as possible reliable markers for HPV infection and lesion development in benign (condyloma acuminatum) and invasive cervical carcinoma (ICC).

We detected the presence of HPV infection in biopsies from 33 women with condyloma acuminatum (M _age=36.5±10.5 years) and 111 ICC patients (M _age=38.9±16.4 years). We performed HPV high-risk (HPV-16, 18, 31, 33, 35, 39, 45, 52, 53, 56, 58, 59, 66, 70) screen with the Sacace Biotechnologies kit (Sacace Biotechnologies, Como, Italy) and nested polymerase chain reaction (PCR) for HPV low-risk (HPV-6, 11) detection (8). A total of 105 control women (M _age=37.5±11.6 years) were selected for the absence of HPV cervical infection and matched with the patient groups for age and risk of infection. All the subjects were recruited at the University Hospital of Ferrara, selected for a documented Caucasian and consistent geographical origin. They were nonsmokers, reported one sexual partner, and were not going through menopause. Clinical assessment and blood sampling were performed during routine clinics, with written informed consent and local ethical board approval. The extraction of DNA was performed on biopsies for HPV identification and on peripheral blood mononuclear cells for HLA-C and KIR typing using protease/SDS digestion, purification by fenol-chloroform followed by ethyl-ether extraction (8).

Due to limited DNA resources, we restricted to the identification of HLA-C allele groups into HLA-C group 1, characterized by Asn at position 80, and group 2, characterized by Lys at position 80 alleles, typed by PCR with specific primers (3). KIR alleles were genotyped by PCR by specific primers (4).

We estimated the frequencies of the HLA/KIR pairs by Arlequin 3.5.1.2 software. We analyzed the distribution of HLA-C alleles, genotypes, and pairs by StatView software package (SAS Institute, Inc., Cary, NC) and GraphPad Prism v6.0 software (Graphpad Software, San Diego, CA). Significance was assumed for p<0.05. Bonferroni correction for multiple comparisons was applied when the p-value was significant.

The HPV analysis reported high-risk HPV types DNA in 100% ICC (111/111) and 9% condyloma (3/33) patients, while low-risk HPV types DNA in 100% condyloma patients (33/33). These data sustain a significant association between high-risk HPV types and ICC development and a strict correlation between condyloma and low-risk HPV types (condyloma vs. ICC=OR 203.5 [95% CI 48.2–858.4]). Three condyloma patients (9%) presented both high- and low-risk HPV infections. This condition was previously demonstrated in other studies, where most condyloma lesions contained multiple HPV types, including types associated with dysplastic epithelial abnormalities (1). The presence of multiple HPV types in a large percentage of condyloma lesions suggests that many individuals acquire additional HPV types at the time of infection with HPV type 6 or 11. Healthy controls presented no positivity for HPV DNA.

First, we considered the possible implication of HLA-C in modifying the risk of HPV infection. We compared HPV negative controls with all HPV positive patients (Test Group) (Table 1). We observed an increased risk of HPV infection in the presence of HLA-C2 allele (OR 2.01 [95% CI 1.4–2.8]) with a higher frequency of both HLA-C2 allele and HLA-C2:-C2 genotypes in HPV positive patients in comparison with HPV negative controls (p _c=0.000048 and 0.0000003 respectively). We then evaluated the effect of HLA-C alleles in high- and low-risk HPV infection (Table 1). We considered only the patients with one type (high or low risk) of HPV infection. We observed an increase in HLA-C1 allele and HLA-C1:-C1 genotype in high-risk HPV-infected patients in comparison with low-risk HPV-infected patients (p _c=0.0019 and 0.033 respectively). These data suggest that the presence of HLA-C1 alleles could facilitate high-risk HPV in comparison with low-risk HPV infection (OR 3.1 [95% CI 1.6–5.98]). We confirmed the results, comparing ICC patients with condyloma patients (Table 2). We observed an increase of HLA-C1 alleles in ICC patients (49%), characterized by high-risk HPV infection, in comparison with condyloma (30%) patients (p _c=0.032).

Table 1.

Allelic, Genotypic, and Pair Frequencies of HLA-C and KIR in Control and Test Groups

	Control group HPV − (150)	Test group HPV + (144)	p _c ^a	Test group HPV + low risk (30)	Test group HPV + high risk (111)	p _c ^b
Allele n (%)
HLA-C1	180 (60)	123 (42)	0.000048^c	14 (23)	08 (49)^c	0.0019^c
HLA-C2	120 (40)	165 (58)		46 (77)	114 (51)
Total	300	288		60	222
Genotype n (%)
HLA-C1:C1	53 (35)	40 (28)		3 (18)	37 (33)
HLA-C1:C2	74 (49)	43 (30)	0.000003^c	8 (24)	34 (31)	0.033^c
HLA-C2:C2	23 (16)	61 (42)		19 (42)	40 (36)
Total	150	144		30	111
Pair n (%)
HLA-C1/KIR2DL2	87 (29)	52 (18)		6 (10)	48 (22)
HLA-C1/KIR2DL3	93 (31)	71 (25)	<0.0001^d	8 (13)	60 (27)
HLA-C2/KIR2DL2	102 (34)	70 (24)		10 (17)	56 (25)	0.000051^d
HLA-C2/KIR2DL3	18 (6)	95 (33)		36 (60)	58 (26)

Control group HPV− vs. test group HPV+.

Test group HPV+ low risk vs. test group HPV+ high risk.

Fisher's exact test.

Chi square test.

p _c, corrected p-value; HPV, human papillomavirus.

Table 2.

Allelic, Genotypic, and Pair Frequencies of HLA-C and KIR in Invasive Cervical Cancer and Condyloma Patients

	ICC (111)	Condyloma (33)	p_c
Allele n (%)
HLA-C1	108 (49)	20 (30)
HLA-C2	114 (51)	46 (70)	0.032^a
Total	222	66
Genotype n (%)
HLA-C1:C1	37 (33)	6 (18)
HLA-C1:C2	34 (31)	8 (24)	0.24^a
HLA-C2:C2	40 (36)	19 (42)
Total	111	33
Pair n (%)
HLA-C1/KIR2DL2	48 (22)	12 (18)
HLA-C1/KIR2DL3	60 (27)	8 (13)	0.00018^a
HLA-C2/KIR2DL2	56 (25)	10 (14)
HLA-C2/KIR2DL3	58 (26)	36 (55)

Fisher's exact test.

ICC, invasive cervical cancer.

Taken together, these results support an implication of HLA-C alleles in HPV infection and lesion development.

Since HLA-C molecules are ligands for KIR2DL2, KIR2DL3, and KIR2DL1 NK receptors, we evaluated the frequencies of these inhibitory receptors. KIR2DL2 and/or KIR2DL3 receptors react with HLA-C1 epitope of HLA-C molecules; KIR2DL2 recognizes C2 in addition to C1, while KIR2DL1 strongly reacts with C2.

Since KIR2DL1 gene was present in the majority of the subject tested (98.4%; www.allelefrequencies.net/kir6006a.asp?kir_positives=on&kir_allele=2DL1), the calculations for this gene were not possible. We subdivided the subjects according to the presence or absence of KIR2DL2 or KIR2DL3 alleles, and analyzed the distribution of the corresponding HLA-C/KIR pairs. We reported a significant difference in the distribution of HLA-C/KIR pairs between control subjects and HPV positive individuals (p _c<0.0001; Table 1). In particular, a decrease in HLA-C2/KIR2DL2 pair was observed in HPV-positive subjects (24%) in comparison with controls (34%; OR 0.54 [95% CI 0.36–0.82]). When we subdivided HPV positive patients on the basis of low- and high-risk HPV types, we observed a significant difference in pair distribution (p _c=0.000051; Table 1). In particular, an increase in HLA-C1/KIR2DL2 and HLA-C1/KIR2DL3 pairs was observed in high-risk HPV in comparison with low-risk HPV infection (OR 3.05 [95% CI: 1.2–8.04]; OR 3.24 [95% CI 1.33–7.89], respectively).

We then evaluated the role of HLA-C/KIR pairs in the development of neoplastic (ICC) and benign (condyloma) lesions. We observed a significant difference between ICC and condyloma subjects (p _c=0.00054; Table 2). In particular, an increase in HLA-C1/KIR2DL2 and HLA-C1/KIR2DL3 pairs was observed in ICC in comparison with condyloma patients (OR 1.33 [95% CI 0.6–2.97]; OR 3.68 [95% CI 1.53–8.86], respectively).

These results suggest HLA-C1/KIR2DL2 and HLA-C1/KIR2DL3 pairs as risk factors for HPV high-risk infection, while the increased frequency of HLA-C2 alleles in HPV-positive patients in comparison with controls did not correspond to a similar increase in HLA-C2/KIR2DL2 pair.

In summary, our results support the involvement of HLA-C alleles in HPV infection and lesion development. Although the presence of HLA-C group 2 alleles leads to an increased risk of HPV infection, the absence of an increased frequency of HLA-C2/KIR2DL2 pair suggests no direct effect on NK cell activation control. On the other hand, HLA-C group 1 antigens are associated with high-risk HPV infection and neoplastic lesion development. Interestingly, the presence of the high inhibitory HLA-C1/KIR2DL2 and HLA-C1/KIR2DL3 interactions has a risk effect toward high-risk HPV infection and neoplastic lesion development. The increase in HLA-C1/KIR2DL2 and HLA-C1/KIR2DL3 pair frequencies in high-risk HPV-infected patients suggests a direct effect on NK cells. In particular, we hypothesize that HLA-C1/KIR2DL2 and HLA-C1/KIR2DL3 presence could maintain NK cell inhibition toward HPV infection sustaining the high-risk HPV-associated transformation and the development of pre-neoplastic lesions (9). The absence of such a strong interaction between HLA-C/KIR pairs and low-risk HPV infection could be due to the lower selective effect of the low-risk HPV infection that leads only to benign lesions.

Recent but few studies suggested a relationship between HLA-C alleles and HPV-related cervical diseases. Four case-control studies have previously examined the effects of variation at the HLA and KIR locus in women with CIN (2,5,6,10,11). In particular, Wang et al. (11) and Martin et al. (6) found that specific HLA ligands for inhibitory KIR were associated with modified risk of developing cervical neoplasia. Song et al. (10) found HLA-C*0303 to confer susceptibility to HPV-related cervical disease, whereas HLA-C*01 was protective against HPV-related cervical disease. Our data are in agreement with the previous results on HLA-C group 1 allele involvement in HPV-associated lesion development (6), and are unique in that they were from an Italian population.

In conclusion, our results strongly implicate HLA-C group 1 alleles in combination with KIR2DL2 and KIR2DL3 receptors as major determinants in high-risk HPV infection and neoplastic lesion development. Of course, these data need to be confirmed in a larger and prospective study and with functional experiments, but they appear to suggest HLA-C and KIR allele analysis as risk markers to be considered in the evaluation of HPV infection risk.

Footnotes

Acknowledgments

We thank Iva Pivanti for skillful technical assistance, Dr. Eleonora Zavatti for statistical assistance, and Linda Marie Sartor for revision of the English language. This work was supported by grants from Fondi Ateneo per la Ricerca (FAR) Università di Ferrara.

Author Disclosure Statement

No competing financial interests exist.

References

Brown

, Schroeder

, Bryan

, Stoler

, abd Fife

. Detection of multiple human papillomavirus types in condylomata acuminata lesions from otherwise healthy and immunosuppressed patients. Clin Microbiol, 1999; 37:3316–3322.

Carrington

, Wang

, Martin

, et al. Hierarchy of resistance to cervical neoplasia mediated by combinations of killer immunoglobulin-like receptor and human leukocyte antigen loci. J Exp Med, 2005; 201:1069–1075.

Chainonthee

, Bottcher

, Gagne

, Fussel

, Bignon

, and Wassmuth

. Improved KIR gene and HLA-C KIR ligand sequence-specific primer polymerase chain reaction genotyping using whole genome amplification. Tissue Antigens, 2010; 76:135–143.

, Gjertson

, Reed

, and Rajalingam

. Receptor-ligand analyses define minimal killer cell Ig-like receptor (KIR) in humans. Immunogenetics, 2007; 59:1–15.

Madeleine

, Johnson

, Smith

, et al. Comprehensive analysis of HLA-A, HLA-B, HLA-C, HLA-DRB1, and HLA-DQB1 loci and squamous cell cervical cancer risk. Cancer Res, 2008; 68:3532–3539.

Martin

, Borecki

, Zhang

, et al. HLA-Cw group 1 ligands for KIR increase susceptibility to invasive cervical cancer. Immunogenetics, 2010; 62:761–765.

Rizzo

, Gentili

, Casetta

, et al. Altered natural killer cells' response to herpes virus infection in multiple sclerosis involves KIR2DL2 expression. J Neuroimmunol, 2012; 251:55–64.

Rotola

, Costa

, Monini

, et al. Impact of sexual habits on the clinical evaluation of male HPV infection. Eur J Epidemiol, 1994; 10:373–380.

Sasagawa

, Takagi

, and Makinoda

. Immune responses against human papillomavirus (HPV) infection and evasion of host defense in cervical cancer. J Infect Chemother, 2012; 18:807–815.

10.

Song

, Lee

, Kim

, et al. Association of KIR genes and HLA-C alleles with HPV-related uterine cervical disease in Korean women. Tissue Antigens, 2013; 81:164–170.

11.

Wang

, Hildesheim

, Gao

, et al. Comprehensive analysis of human leukocyte antigen class I alleles and cervical neoplasia in 3 epidemiologic studies. J Infect Dis, 2002; 186:598–605.

12.

zur Hausen

. Papillomavirus infections: a major cause of human cancers. Biochim Biophys Acta, 1996; 1288: F55–78.