Loss of Epidermal Langerhans Cells Occurs in Human Papillomavirus α, γ, and μ but Not β Genus Infections

ORIGINAL ARTICLE

Loss of Epidermal Langerhans Cells Occurs in Human Papillomavirus a, c, and l but Not b Genus Infections Cheng Mee Leong1, John Doorbar2, Ingo Nindl3, Han-Seung Yoon1 and Merilyn H. Hibma1 Human papillomaviruses (HPVs), which are contained in the a, b, g, m, and n genera, differ in their oncogenic potential and their tropism for cutaneous or mucosal epidermis. Langerhans cells (LC), the only epidermal professional antigen-presenting cells, are readily detected in normal mucosal and cutaneous epithelium. The aim of this study is to determine whether LC loss, which has been reported for HPV16, occurs in other HPV genera and establish its significance in viral pathology. We found that, as for HPV16, LCs were reduced in lesions infected with high-risk mucosal (a7 and a9 species) and low-risk cutaneous (g and m) types. Lesions infected with a10 low-risk genital types had reduced LC but contained epidermal LC patches, coincident with dermislocalized regulatory T cells (T-regs). In contrast to other genera, LCs were common in the epidermis, and T-regs occupied the dermis of the potentially high-risk cutaneous b-HPV type infected lesions. Therefore, LC loss in the infected lesions occurred irrespective of tropism or oncogenic potential of the HPV type. LC depletion in the HPV-infected epidermis may create an environment that is permissive for viral persistence and in HPV lesions in which LCs are found, the presence of typically immunosuppressive T-regs may compensate for their continued presence. Journal of Investigative Dermatology (2010) 130, 472–480; doi:10.1038/jid.2009.266; published online 17 September 2009

INTRODUCTION Langerhans cells (LCs) are the primary antigen-presenting cells found in the epidermis. The classical view of their role is one of antigen uptake in the epidermis, and migration through the dermal lymphatics to the lymphoid organs, where they present antigen to lymphocytes that then home back to the tissue to carry out their effector function. This is considered of critical importance in the defense against pathogens that are present only in the epidermis. Human papillomaviruses (HPVs) are one of a small group of viruses whose infection is confined to the epidermis where LCs are resident. The over 100 HPV genotypes are subdivided into five genera, a, b, g, m, and n, based on the DNA sequence of the major capsid protein L1 (de Villiers et al., 2004). There is diversity in the pathology of the HPV types across the genera and species, in particular in relation to the nature of the epidermis infected and the oncogenic potential of the viral type. The a genus is heterogeneous, containing the majority of the HPV types that cause overt lesions. It includes a7 and a9 1

Departments of Microbiology and Immunology and Pathology, University of Otago, Dunedin, New Zealand; 2National Institute for Medical Research, Mill Hill, London, UK and 3DKFZ-Charite Cooperation, Viral Skin Carcinogenesis, German Cancer Research Center, Heidelberg, Germany Correspondence: Merilyn H. Hibma, Departments of Microbiology and Immunology, University of Otago, PO Box 56, Dunedin, New Zealand. E-mail: [email protected] Abbreviations: HPV, human papillomavirus; LC, Langerhans cell; T-regs, regulatory T cells Received 1 May 2009; revised 15 July 2009; accepted 17 July 2009; published online 17 September 2009

472

Journal of Investigative Dermatology (2010), Volume 130

species, containing the high-risk, cancer-causing mucosal HPV types 16, 18, 31, and 45. Other a species include the a10 species, which contains the low-risk mucocutaneous genital types 6 and 11, and the a4 species containing types 2, 27, and 57, which are most frequently associated with common skin warts in the European population (Rubben et al., 1997). Other HPV containing genera are less heterogeneous. DNA from the b types can be found at low copy number in cutaneous epidermis of immune competent individuals (Weissenborn et al., 2005), but in high copy number in individuals with Epidermodysplasia verruciformis and those who are immune compromised (Dell’Oste et al., 2009). The b types cause flat lesions in Epidermodysplasia verruciformis patients and b DNA can be found in cancerous lesions such as actinic keratoses, but overt HPV lesions are rare in immune competent individuals (Pfister, 2003). The g and m types cause proliferative cutaneous lesions in humans, but differ from each other in that the g genus, like the b genus, is missing the E5 open reading frame. Many HPV types cause persistent infections, lasting for months to years. Cutaneous types such as types 1 and 2, or mucocutaneous types, such as types 6 and 11, cause lesions that can become sizable, with large amounts of infectious particles being produced. The ability of many HPV types to cause persistent infections with significant quantities of viral antigens being produced has been attributed to viral immune evasion mechanisms that prevent immune recognition or dampen the immune response. The restriction of HPV antigen expression to the epidermis implicates the LC as the sole professional antigen-presenting population positioned to take up HPV antigen. We have & 2010 The Society for Investigative Dermatology

CM Leong et al. HPV Types and Langerhans cells

previously reported that LC numbers are significantly reduced in a9 HPV type 16 lesions (Matthews et al., 2003). In this report, we show that LCs were reduced in both cutaneous (a4, g, and m) and mucosal (a7 and a9) HPV-infected epithelium. Only HPV types from the a10 and b genus contained LC, which, in the case of a10 types, was confined to patches. In b lesions, the frequency of LC was comparable to normal cutaneous epidermis and was incongruous with LC depletion contributing to the ability of the virus to avoid detection by the immune system. However, in every case where LCs were present, regulatory T cells (T-regs) were in close proximity. We conclude that LC depletion contributes to the ability of the virus to produce virus particles and persist in the presence of a hostile host immune system and that when LC are present in an HPV-infected lesion, other immune suppressive mechanisms, such as infiltrates of T-regs contribute to the suppressive environment. RESULTS We have previously described reduced LC frequencies in HPV16-positive low-grade lesions (Matthews et al., 2003) but it is not known whether LC depletion is also a function of other high-risk types included in the a genus. To look at this, we examined a number of low-grade lesions infected with other high-risk HPV types of the a7 and a9 species, including types 18, 31, 33, 35, and 45. The selected tissues were all positive for the expression of the early/late viral protein-E4, which was taken to indicate that productive stages of the viral life cycle are supported, and viral genomes are being amplified in preparation for packaging. LCs were stained with an antibody specific for CD1a, LC being the only CD1apositive cells normally found in the epidermis. CD1a staining of tissues infected with HPV18 or 31, representative of the staining observed in the high-risk HPV types tested, is shown in Figure 1. Overall, we found that LCs were rarely detected in mucosal lesions infected with these high-risk types compared with their frequent observation in normal cervical tissue. Langerhans cells per mm2 of epidermis in the a7 and Normal cervical cer(17)

a

a9 tissues was determined and all of the lesions tested showed around a threefold reduction in LC compared with normal cutaneous epidermis (Figure 2). LC frequencies from both the a7 and the a9 species had significantly lower LC (P ¼ 0.02 and P ¼ 0.02 respectively; Unpaired t-test with Welch’s correction) compared with those from normal epidermis. This suggests that viral regulation of LC loss is broadly conserved amongst the most common high-risk mucosa infecting HPV types of the a7 and a9 species. We wished to determine whether LC loss also occurred in low-risk HPV types of the a10 species. The a10 HPV types are evolutionarily separated from the a7 and a9 species, retaining between 60 and 70% similarity in the L1 nucleotide sequence (de Villiers et al., 2004). HPV6 and 11 are prevalent HPV types of the a10 genus, causing condylomata acuminata. These infections can present as highly proliferative, florid lesions of the cutaneous or mucosal epithelium of the genital region. To test whether LCs were lost from a10-infected epidermis, seven condylomata acuminata (three from cutaneous and four from mucosal sites) obtained from different individuals were stained for CD1a-positive LC. LCs were significantly reduced overall in the a10-infected tissues compared with the respective normal tissues, regardless of the site from which they were obtained (cutaneous, P ¼ 0.0003; mucosal, P ¼ 0.014); however, the pattern of staining differed from the a7- and a9-infected tissues (Figure 3). LCs were more frequently observed in the a10 compared with a7- and a9-infected tissues, and when present tended to be concentrated in patches of infected epidermis, with other areas devoid of LC. Overall, LC numbers were reduced around twofold on average in the mucosal and around eightfold in the cutaneous condyloma lesions compared with normal mucosal or cutaneous epidermis (Figure 2). As LC numbers were reduced in the skin infected with highrisk a7 and a9 types and the low-risk a10 types, we conclude that there is no correlation between the oncogenic potential of the viral type and LC loss. In a10 lesions, LC could be found in patches at frequencies similar to that of normal cutaneous epidermis. HPV31 cer(6)

HPV18 cer(1)

c

b E D

E

E

D D

D

Figure 1. Reduced LC density in a7 and a9 high-risk HPV-infected cervical tissue. Normal cervical tissue and HPV-infected lesions were stained with an antibody to CD1a. (a) Normal cervical tissue, cer(17) showing numerous LC in the basal and parabasal layers (inset in upper-right corner shows enlargement of CD1a-positive cells); (b) HPV18-infected tissue, cer(1), with no LC; (c) HPV31-infected tissue, cer(6), with the few LC indicated by arrows. E, epidermis; D, dermis. Broad dashed line indicates the interface between the epidermis and dermis and narrow dashed line indicates the surface of the epidermis. Bar ¼ 50 mM.

www.jidonline.org

473

CM Leong et al. HPV Types and Langerhans cells

Mucosal lesions

a

*

Langerhans cells per mm2

600

*

*

400

200

10

9

α

α

7 α

N m or m uc a os l al

0

Cutaneous lesions

b

***

Langerhans cells per mm2

*** ***

2500

***

2000 1500 1000 500

10 α

4 α

μ

γ

β

cu No ta rm ne al ou s

0

Figure 2. LC frequency in epidermis infected with HPV a, b, c, and l types. LC in HPV-infected mucosal (a) or cutaneous (b) lesions were enumerated and the number of Langerhans cells per mm2 is presented. Dots represent individual samples. Lines indicate the mean of the samples for each group. There is a significant difference in the mean number of Langerhans cells per mm2 between the mucosal groups (One-way analysis of variance, Po0.05) and between the cutaneous groups (Po0.001). Pair-wise comparisons were carried out between the normal group and the test (excluding a4 and m groups) using the unpaired t-test with Welch’s correction. Statistical differences are indicated by asterisks: *Po0.05; ***Po0.001. There was no significant difference between the b group and normal controls (P ¼ 0.89).

Dermal leukocyte infiltration was observed in many of the a10-infected tissues, indicative of an immunologically active microenvironment that is at odds with a persistent and active viral infection, as evidenced pathologically by the highly proliferative epidermis. One explanation for the presence of leukocyte infiltrates in the absence of apparent regression might involve T-regs, which have been described in condylomata (Cao et al., 2008) and could exert a suppressive effect on HPV-specific effector T cells. To establish whether T-regs were localized in or around the patches of LC found in the condylomata, sequential tissues were stained for the leukocyte common antigen, CD45, FoxP3, a specific marker for CD25 þ T-reg cells, and for LC using CD1a. Representative examples of the staining of mucosal and cutaneous a10infected epithelium compared with normal cutaneous and 474

Journal of Investigative Dermatology (2010), Volume 130

mucosal epithelium are shown in Figure 4. Consistently, patches of LC in the epidermis were localized to areas that contained CD45-positive infiltrates in the proximal dermis (Figure 4). FoxP3-positive cells were readily identified in the dermal infiltrates and were also occasionally identified in the epidermis. We were unable to identify regions of tissue that were LC positive but devoid of FoxP3-positive cells either in the dermis or epidermis. In contrast, regions of tissue without LC but containing CD45-positive and FoxP3-positive cells in the dermis or epidermis could be found (data not shown). These data support the presence of T-regs in condylomata, which may, in conjunction with the overall reduction in LC density, contribute to immune evasion by the a10 species. The HPV types that typically cause cutaneous lesions from the a, g, and m genera are generally considered to be benign, proliferative, and productive. In contrast, the b types are generally associated with asymptomatic infections of cutaneous epidermis in immune competent individuals and overt lesions in immunocompromised individuals. There is increasing evidence for an association between infection with b HPV types and squamous cell carcinoma (Karagas et al., 2006) and the risk of progression to malignancy is elevated in immune suppressed individuals and those with Epidermodysplasia verruciformis. To our knowledge, the effects of lowrisk cutaneous HPV from the a4, g, and m genera on LC number at the site of infection has not previously been reported, perhaps in part due to the difficulty in obtaining these types of tissues, as surgical excision is no longer widely used for the removal of cutaneous warts. We were able to obtain a limited number of archival tissues from the a4 (4), b (3), g (3), and m (2) genera that were, with the exception of the b infected tissues, representative of the nature of lesions caused by types within each of the genera. The b lesions were unusual in that they were overt lesions from immune competent individuals, which are rarely found. Consistent with the other a types tested here, CD1a staining of verruca vulgaris lesions caused by HPV type 2, a commonly found type of the a4 genus, had few LC in the epidermis (Figure 5). This tissue had a greater than tenfold reduction in LC (Po0.0001) when compared with normal epidermis (Figure 2), as LC were generally not detected. Similarly, tissues infected with types 4 or 65, belonging to the g genus, or with types 1 or 63 of the m genus, also had fewer detectable LC after CDla staining compared with normal epidermis (g: Po0.0001; m: Po0.0001) (Figure 5). In contrast, all the cutaneous tissues infected with only b HPV types (Figure 6) had equivalent or higher numbers of LC compared with cutaneous epidermis (P ¼ 0.89). The difference in LC density in b lesions was marked and can be seen in Figure 6. An additional tissue infected with a b type, HPV76, was co-infected with the a4 type 57b and the m type 1. This lesion retained reduced LC density characteristics consistent with the a and m genera (Figure 6g), rather than the higher density observed with the other b-infected tissues. We questioned how proliferative lesions caused by these b types might be able to persist in immune competent individuals. On the basis of our and other’s observations of T-regs in condylomata, we stained the b-containing tissues

CM Leong et al. HPV Types and Langerhans cells

gen(8) low

gen(8) high

a

Normal vulval cut(19)

b

c

D

E

E E

D

d

e

f

E

E E

D D

D gen(1) low

gen(1) high

Normal cervical cer(13)

Figure 3. LC density is patchy but is reduced overall in condylomata acuminata lesions. Two regions of two tissues, gen(8) from a vulval lesion and gen(1) from a cervical lesion, stained for CD1a are shown; (a) and (d) are regions devoid of LC for tissues gen(8) and gen(1), respectively, and (b) and (e) are regions of gen(8) and gen(1) with LC present. Normal cutaneous vulval epidermis (c), cut(19), and normal mucosal cervical epidermis (f), cer(13), are also shown. E, epidermis; D, dermis. Broad dashed line indicates the interface between the epidermis and dermis and narrow dashed line indicates the surface of the epidermis. Bar ¼ 50 mM.

with FoxP3 to detect T-regs. We found that the dermis beneath b-infected epidermis contained T-regs, which were not detected in the dermis of the co-infected or normal cutaneous epidermis (Figure 6). DISCUSSION It is widely accepted that because LC are the only professional antigen-presenting cells in the epidermis they will contribute, either directly or indirectly, to the initiation of an immune response to epidermis-infecting pathogens. In this study, we show that LC numbers are markedly reduced by 2- to 175-fold in tissue infected with members of the a, g, and m genera. Regulation of LC number in tissue was not related to the oncogenic potential of the virus as tissue from high-risk types, such as 18 and 45 and low-risk types, such as 1 and 2, and the condylomata acuminata lesions had reduced LC density at the site of infection. Similarly, HPV types seemed competent to regulate LC abundance in cutaneous (for example, HPV1) or mucosal tissue (for example, HPV18). Using only tissues infected with b types tested here, we were able to show no reduction in LC density. We acknowledge that the size of the b group tested here is small, reflecting the difficulty in obtaining overt lesions caused by b types in immune competent individuals. Consequently, this result may not be representative of all b types. However, these data

are supported by a previous report of minimal change in LC density in b lesions, albeit from Epidermodysplasia verruciformis patients (Cooper et al., 1990). Unlike the other HPV types tested here, b HPV types are typically associated with asymptomatic infections that produce only a low copy number of viral genomes in immune competent individuals (Weissenborn et al., 2005). The b tissues that were tested in this study were uncharacteristic, as they showed evidence of skin thickening compared with normal epidermis. The presence of LC in these tissues indicates that HPV can coexist with resident LC; however, we propose that this is achieved by viral manipulation of the immune response resulting in the recruitment of suppressive T-regs that were readily identified in the dermis below the infected site. As b types are typically asymptomatic in immune competent individuals, we speculate that b types may have evolved to be able to maintain an infection in the presence of functional and active LC because they replicate at low copy number and with limited viral antigen expression. However, the maintenance of larger productive lesions such as condylomata involves the additional immunosuppressive effects of dermal T-regs. We have previously reported for HPV16 (Matthews et al., 2003) and further observed in this study that in HPV types in which a reduction in LC density is seen, and normal skin is www.jidonline.org

475

CM Leong et al. HPV Types and Langerhans cells

CD1a

CD45

a

FoxP3

b

c

gen(2)

E

E

E D D

D

E

d

e

f E

E cer(14)

E

D

D

h

g

D

i

E

gen(8)

E E

D

D

D

j

k

l

E

E

cut(20)

E

D

D

D

Figure 4. Condyloma-derived tissue with patchy LC staining has T-regs present in the adjacent epidermis. Sequential sections of condyloma tissues from a10-infected cervical, gen(2), (a–c), normal cervical, cer(14), (d–f) and a10-infected cutaneous, gen(8), (g–i) and normal vulval, cut(20), (j–l) tissue. Tissues were stained with antibodies to CD1a (a, d, g, and j), CD45 (b, e, h, and k), and FoxP3 (c, f, i, and l). LC-, CD45-, and FoxP3-positive regions of tissue are shown for gen(2) and gen(8). E, epidermis; D, dermis. Broad dashed line indicates the interface between the epidermis and dermis. Bar ¼ 50 mM.

attached to the lesion, LC depletion does not occur in the surrounding uninfected tissue. From this, we conclude that HPV only affects the immediate microenvironment. The epidermis is normally repopulated with LC by TGF-b1-driven 476

Journal of Investigative Dermatology (2010), Volume 130

proliferation within the epidermis (Merad et al., 2002) and, under inflammatory conditions, by chemokine and TGFb1signaled infiltration and differentiation of CD34 þ hematopoietic progenitor cells (Larregina et al., 2001). HPV might

CM Leong et al. HPV Types and Langerhans cells

b

E EE D

c

α4 - cut(1)

Normal cutaneous cut(11)

a

d E

γ - cut(8)

μ - cut(10)

E

D

D

Figure 5. LCs are depleted in cutaneous tissue infected with a, c , or l HPV. Normal cutaneous (a) tissue staining with CD1a displaying LC throughout the epidermis (inset in lower left shows enlargement of CD1a-positive cells) and (b) a4 (HPV2, cut(1)), (c) g (HPV4 and 65, cut(8)), and (d) m (HPV63, cut(10)) tissue stained for CD1a showing few detectable CD1a-positive LC. E, epidermis; D, dermis. Broad dashed line indicates the interface between the epidermis and dermis and narrow dashed line indicates the surface of the epidermis. Bar ¼ 50 mM.

affect one or both of these repopulation pathways to reduce LC density in the skin. Several other virally mediated mechanisms contribute to the loss of LC in HPV-infected epidermis. These include regulation of cell adhesion molecules, such as the HPV16 E6and E7-mediated downregulation of E-cadherin, the cell surface molecule that adheres keratinocytes and LC together and is important in LC retention in the epidermis (Matthews et al., 2003; Caberg et al., 2008), and the regulation of chemokines, such as HPV16 E6 and E7 downregulation of MIP3a/CCL20, a molecule important in LC infiltration into the epidermis (Guess and McCance, 2005; Caberg et al., 2009). It is interesting to note that in each case these regulatory functions are attributed to the HPV protooncogenes, the primary function of which is to permit cell cycle progression in differentiated cells of the epidermis. The apparent inability of the b types in the lesions tested here to regulate LC may reside in functional differences in E6 and E7. Some of the tissues infected with a10 HPV types, despite being large with highly thickened epidermis and having reduced LC overall, contained patches of tissue in which LC could be clearly identified. We found clear evidence of T-regs also in these lesions in which LCs were present.

Cao et al. (2008) also have reported T-regs in condylomata, which have a suppressive function in culture. Jiang et al. (2007) reported that the disruption of E-cadherin adhesion on LC promotes a T-reg response; therefore, HPV regulation of LC maturation, in addition to LC migration and proliferation, may contribute to the development of a suppressor rather than an effector response. T-regs are more likely to be present in large condylomata (Cao et al., 2008), suggesting that HPV antigen load is the trigger for LC recruitment and lymphocyte infiltration once a lesion has become sizable. The suppressive environment that is created by the presence of T-regs would aid viral persistence, although ultimately these lesions will also regress. Although HPV infections can persist for months or years, the majority of lesions do eventually regress in immune competent individuals. As HPV antigens are only expressed in the epidermis, a role for LC in disease regression is implicated, which may be initiated by lesion size, abundance, or an inflammatory event, such as damage to the lesion. It has been shown that HPV VLP can be taken up by LC; however, their immune function is disrupted by viral interference of the PI3K pathway (Fausch et al., 2005). Cross presentation of viral antigens other than the capsid would circumvent this interference and T-cell responses to other viral antigens are implicated in protective immunity against HPV16 (de Jong et al., 2004; Dillon et al., 2007). Whether the LC themselves present antigen or whether viral antigens are transferred from LC to dermal DC is unknown. Initiation of an immune response to HPV is a critical event in natural regression that may involve functionally competent LC. The data presented in this study show that the majority of persistent viral lesions contain few LC and that when a lesion contains normal LC numbers, T-regs are present in the dermis beneath the infection. This suggests that LC regulation and T-reg recruitment are significant events in viral manipulation of the epidermis that contribute to persistence. Developing new therapies that trigger the infiltration of LC into HPV lesions and/or deplete T-regs from the dermis beneath the site of infection may improve outcomes for HPV-infected individuals in the future. MATERIALS AND METHODS Tissue specimens and staining Archival biopsy material (Table 1) was obtained without reference to patient details and with local ethical approval. E4 staining of the HPV16-positive cervical tissues has previously been reported (Matthews et al., 2003). For other mucosal HPV types (HPV18, 33, 35, and 45) and the b types, a panel of rabbit anti-E4 antibodies raised against the relevant maltose-binding protein-E4 fusion proteins were used to confirm productive infection. DNA isolation and HPV DNA typing of verruca vulgaris containing b-PV types were detected as described by Brink et al. (2005). The sources and HPV typing of the remaining tissue has been described previously (Peh et al., 2002; Middleton et al., 2003). Formalin-fixed, paraffinembedded tissues were cut at 4-mm thickness, mounted on glass slides, and air-dried at 37 1C overnight before staining. Sections were deparaffinized in xylene and graded alcohols. Antigen retrieval was carried out by microwaving sections at 800 W in Tris-HCl pH 9.5 www.jidonline.org

477

CM Leong et al. HPV Types and Langerhans cells

β cut(4)

β cut(2)

a

c

D

β+μ+α4 cut(5)

β cut(3)

e

Normal cutaneous cut(13)

g

i

E

E

CD1a

E D

E

E E

D

D

D

D

b

d

f

E

h

j

D

E

FoxP3

D

E

E D

E

D D

Figure 6. LC density is not reduced and T-regs are more frequent in b lesions. CD1a and FoxP3 staining of lesions infected with b HPV types have high numbers of LC (a, c, and e) in the epidermis, comparable to normal cutaneous epidermis (i) and T-regs (indicated by arrows) are frequently observed in b-infected dermis (b, d, and f) but not detected in normal epithelium (j). LC density was reduced (g) and T-regs were absent (h) from epithelium co-infected with b, m, and a4-HPVs. E, epidermis; D, dermis. Broad dashed line indicates the interface between the epidermis and dermis and narrow dashed line indicates the surface of the epidermis. Bar ¼ 50 mM.

Table 1. Summary of clinical data of the 49 patients studied Patient no.

Anatomical site/epidermis type

Genus

HPV type

Histological diagnosis

cut(1)

Cutaneous

a/4

2

Verruca vulgaris

cut(16)

Cutaneous

a/4

2

Verruca vulgaris

cut(17)

Cutaneous

a/4

2

Verruca vulgaris

cut(18)

Cutaneous

a/4

2

Verruca vulgaris

cer(1)

Cervical/mucosal

a/7

18

CIN1

cer(2)

Cervical/mucosal

a/7

18

CIN1

cer(3)

Cervical/mucosal

a/7

45

CIN1

cer(4)

Cervical/mucosal

a/7

45

CIN1

cer(5)

Cervical/mucosal

a/7, a/9

18/35

CIN1

cer(6)

Cervical/mucosal

a/9

31

CIN1

cer(7)

Cervical/mucosal

a/9

31

CIN1

cer(8)

Cervical/mucosal

a/9

16

CIN1

cer(9)

Cervical/mucosal

a/9

16

CIN1

cer(10)

Cervical/mucosal

a/9

16/33

CIN1

cer(11)

Cervical/mucosal

a/9

33

CIN1

gen(1)

Cervical/mucosal

a/10

11

Condyloma acuminata

gen(2)

Cervical/mucosal

a/10

11

Condyloma acuminata

gen(3)

Cervical/mucosal

a/10

6

Condyloma acuminata

gen(4)

Vulval/perineum/cutaneous

a/10

6

Condyloma acuminata

gen(5)

Perianal/cutaneous

a/10

11

Condyloma acuminata Table 1 continued on following page

478

Journal of Investigative Dermatology (2010), Volume 130

CM Leong et al. HPV Types and Langerhans cells

Table 1. Continued Patient no.

Anatomical site/epidermis type

Genus

HPV type

Histological diagnosis

gen(7)

Cervical/mucosal

a/10

6

Condyloma acuminata

gen(8)

Vulval/perianal/cutaneous

a/10

6

Condyloma acuminata

cut(2)

Shoulder/cutaneous

b1

12

Verruca vulgaris

cut(3)

Shoulder/cutaneous

b1,b2

21/37

Verruca vulgaris

cut(4)

Forearm/cutaneous

b1

20

Verruca vulgaris

cut(5)

Ear/cutaneous

b/3, m1, a4

76, 1, 57b

Verruca vulgaris

cut(6)

Cutaneous

g/1

65

Verruca vulgaris

cut(7)

Cutaneous

g/1

4

Verruca vulgaris

cut(8)

Cutaneous

g/1

4/65

Verruca vulgaris

cut(9)

Cutaneous

m/1

1

Verruca vulgaris

cut(10)

Cutaneous

m/2

63

Verruca vulgaris

cut(11)

Cutaneous

Normal

cut(12)

Cutaneous

Normal

cut(13)

Cutaneous

Normal

cut(14)

Cutaneous

Normal

cut(15)

Cutaneous

Normal

cut(19)

Vulval/cutaneous

Normal

cut(20)

Vulval/cutaneous

Normal

cer(12)

Cervical/mucosal

Normal

cer(13)

Cervical/mucosal

Normal

cer(14)

Cervical/mucosal

Normal

cer(15)

Cervical/mucosal

Normal

cer(16)

Cervical/mucosal

Normal

cer(17)

Cervical/mucosal

Normal

cer(18)

Cervical/mucosal

Normal

cer(19)

Cervical/mucosal

Normal

cer(20)

Cervical/mucosal

Normal

cut(21)

Cutaneous

Normal

cut(22)

Cutaneous

Normal

HPV, human papillomavirus.

containing 5% urea for 10 minutes, twice in succession. Slides were cooled to RT then washed in dH2O and equilibrated in TBS, pH 7.6. Sections were blocked for 30 minutes at RT in 5% normal sheep serum in TBS. Sections were incubated for 1.5 hours at RT with primary antibodies (aCD1a, Santa Cruz Biotech, Santa Cruz, CA; aFoxP3, Santa Cruz Biotech; aCD45, Dako, Glostrup, Denmark). After washing with TBS (3  5 minutes), sections were incubated with secondary antibody (anti-mouse Alexa 594, Invitrogen, Carlsbad, CA) and the DNA-binding dye, DAPI, for 1 hour at RT. Sections were washed in TBS and then mounted in Slowfade Gold (Invitrogen) mounting medium.

LC enumeration Images of stained epidermis were acquired in TIFF format with the scale marked using an Olympus BX51 fluorescence microscope

and DP Manager software (Olympus, Center Valley, PA). TIFF images were analyzed using Image J software (http://rsbweb.nih.gov/ ij/). The localization of the epidermis was determined using the nuclear DAPI staining as a guide and was traced using the polygon tool. The epidermal area was determined using the ‘‘analyze/ measure’’ function after calibration to the marked scale. For CD1a immunoreactivity, LC with at least two well-identified dendrites were considered positive. The whole specimen was counted for the majority of tissues; however, for large genital tissues, regions of tissue were randomly selected and counted. A minimum of 5 mm2 of epidermis was counted in those cases. LCs were manually tagged using the Image J plug-in ‘‘cell_counter.jar’’. The number of Langerhans cells per cross-sectional mm2 epidermis was calculated by dividing the number of LC by the calculated area. Tissues were evaluated by two people independently. Statistical analysis was www.jidonline.org

479

CM Leong et al. HPV Types and Langerhans cells

carried out using an unpaired t-test with Welch’s correction (does not assume equal variances). CONFLICT OF INTEREST Dr Hibma is an advisory committee member for GSK (NZ). The remaining authors state no conflict of interest.

ACKNOWLEDGMENTS The authors acknowledge the considerable help from many people (clinical scientists, dermatologists, and GU clinicians) in assembling the range of HPV tissues used in this study. In particular, we are thankful for help in obtaining the HPV types that were provided by Catherine Harwood (Centre for Cutaneous Research, London, UK). This research was funded by the Health Research Council of New Zealand and the Cancer Society of New Zealand.

REFERENCES

Dillon S, Sasagawa T, Crawford A, Prestidge J, Inder MK, Jerram J et al. (2007) Resolution of cervical dysplasia is associated with T-cell proliferative responses to human papillomavirus type 16 E2. J Gen Virol 88:803–13 Fausch SC, Fahey LM, Da Silva DM, Kast WM (2005) Human papillomavirus can escape immune recognition through Langerhans cell phosphoinositide 3-kinase activation. J Immunol 174:7172–8 Guess JC, McCance DJ (2005) Decreased migration of Langerhans precursorlike cells in response to human keratinocytes expressing human papillomavirus type 16 E6/E7 is related to reduced macrophage inflammatory protein-3alpha production. J Virol 79:14852–62 Jiang A, Bloom O, Ono S, Cui W, Unternaehrer J, Jiang S et al. (2007) Disruption of E-cadherin-mediated adhesion induces a functionally distinct pathway of dendritic cell maturation. Immunity 27: 610–624 Karagas MR, Nelson HH, Sehr P, Waterboer T, Stukel TA, Andrew A et al. (2006) Human papillomavirus infection and incidence of squamous cell and basal cell carcinomas of the skin. J Natl Cancer Inst 98:389–95

Brink AA, Lloveras B, Nindl I, Heideman DA, Kramer D, Pol R et al. (2005) Development of a general-primer-PCR-reverse-line-blotting system for detection of beta and gamma cutaneous human papillomaviruses. J Clin Microbiol 43:5581–7

Larregina AT, Morelli AE, Spencer LA, Logar AJ, Watkins SC, Thomson AW et al. (2001) Dermal-resident CD14+ cells differentiate into Langerhans cells. Nat Immunol 2:1151–8

Caberg JH, Hubert P, Herman L, Herfs M, Roncarati P, Boniver J et al. (2009) Increased migration of Langerhans cells in response to HPV16 E6 and E7 oncogene silencing: role of CCL20. Cancer Immunol Immunother 58:39–47

Matthews K, Leong CM, Baxter L, Inglis E, Yun K, Backstrom BT et al. (2003) Depletion of Langerhans cells in human papillomavirus type 16-infected skin is associated with E6-mediated down regulation of E-cadherin. J Virol 77:8378–85

Caberg JH, Hubert PM, Begon DY, Herfs MF, Roncarati PJ, Boniver JJ et al. (2008) Silencing of E7 oncogene restores functional E-cadherin expression in human papillomavirus 16-transformed keratinocytes. Carcinogenesis 29:1441–7

Merad M, Manz MG, Karsunky H, Wagers A, Peters W, Charo I et al. (2002) Langerhans cells renew in the skin throughout life under steady-state conditions. Nat Immunol 3:1135–41

Cao Y, Zhao J, Lei Z, Shen S, Liu C, Li D et al. (2008) Local accumulation of FOXP3+ regulatory T cells: evidence for an immune evasion mechanism in patients with large condylomata acuminata. J Immunol 180:7681–6

Middleton K, Peh W, Southern S, Griffin H, Sotlar K, Nakahara T et al. (2003) Organization of human papillomavirus productive cycle during neoplastic progression provides a basis for selection of diagnostic markers. J Virol 77:10186–201

Cooper KD, Androphy EJ, Lowy D, Katz SI (1990) Antigen presentation and T-cell activation in epidermodysplasia verruciformis. J Invest Dermatol 94:769–76

Peh WL, Middleton K, Christensen N, Nicholls P, Egawa K, Sotlar K et al. (2002) Life cycle heterogeneity in animal models of human papillomavirus-associated disease. J Virol 76:10401–16

de Jong A, van Poelgeest MI, van der Hulst JM, Drijfhout JW, Fleuren GJ, Melief CJ et al. (2004) Human papillomavirus type 16-positive cervical cancer is associated with impaired CD4+ T-cell immunity against early antigens E2 and E6. Cancer Res 64:5449–55

Pfister H (2003) Chapter 8: human papillomavirus and skin cancer. J Natl Cancer Inst Monogr 31:52–6

de Villiers EM, Fauquet C, Broker TR, Bernard HU, zur Hausen H (2004) Classification of papillomaviruses. Virology 324:17–27 Dell’Oste V, Azzimonti B, De Andrea M, Mondini M, Zavattaro E, Leigheb G et al. (2009) High beta-HPV DNA loads and strong seroreactivity

480

are present in epidermodysplasia verruciformis. J Invest Dermatol 129:1026–34

Journal of Investigative Dermatology (2010), Volume 130

Rubben A, Kalka K, Spelten B, Grussendorf-Conen EI (1997) Clinical features and age distribution of patients with HPV 2/27/57-induced common warts. Arch Dermatol Res 289:337–40 Weissenborn SJ, Nindl I, Purdie K, Harwood C, Proby C, Breuer J et al. (2005) Human papillomavirus-DNA loads in actinic keratoses exceed those in non-melanoma skin cancers. J Invest Dermatol 125:93–7

Copyright of Journal of Investigative Dermatology is the property of Nature Publishing Group and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.