Impaired HPV-Specic T-Cell Response Characterized with Skewed Cytokine Prole and T-Cell Dysfunction in Juvenile-Onset Recurrent Respiratory Papillomatosis Patients

Immunologic dysfunction is one of the most important mechanisms underlying the initiation and development of Juvenile-onset Recurrent Respiratory Papillomatosis (JORRP). The study aimed to explore whether HPV-specic T-cell response was impaired in JORRP patients. A total of 46 JORRP patients and 93 age- and sex- matched healthy controls were enrolled. The plasma concentrations of various cytokines were measured using the Luminex. The cytokine mRNA proles of PBMCs were quantied with the real-time PCR. The T-cell subsets, HPV-specic T-cell activation, proliferation, apoptosis and expression of replicative senescent T-cell markers CD57 were measured with the ow cytometry. The cytokine secretion of HPV-specic T cells was assessed by the ELISPOT.


Introduction
Recurrent respiratory papillomatosis (RRP) is a benign disease of the upper aero-digestive tract caused by human papillomavirus (HPV) infection, in particular low-risk virus type HPV6 and HPV11 [1,2]. RRP typically presents in a bimodal age pattern occurring either during adult life or early childhood [3]. Adultonset RRP (AORRP) typically emerges in early adulthood, the most vulnerable age window when HPV is readily acquired by sexual transmission. By contrast, juvenile-onset RRP (JORRP) appears in early childhood and usually is transmitted vertically during pregnancy or is acquired at birth from an HPV-infected mother [3]. Of note, JORRP has a more aggressive and recurring clinical course, perhaps in part because infants have smaller airways or because of immune compromise [4]. Surgical excision is the current standard of care for RRP, which is directed at preventing upper airway compromise and improving vocal function while preserving laryngeal tissues [5]. However, there is no cure for RRP so far. Typical JORRP patients require nearly 20 surgical procedures throughout their lifetimes, many of which occur early in the children's lives. During the initial years of the disease, a child is estimated to require slightly more than 4 surgeries per year [6]. Thus, this disease not only affects individual patients signi cantly but also places a large economic burden on their families and society in general [7]. To date, it is unclear why some individuals do not mount a su cient anti-HPV response to the initial HPV6-and HPV11 infection [8].
Researchers found evidence that T cell immune dysfunction may contribute to the development of RRP.
One group observed that IL-4 and IL-10 were dominantly expressed, compared with IFN-γ, in T cells that in ltrate papilloma of RRP patients and this imbalance correlated with disease severity [9]. Another study proposed that RRP patients have more circulating CD4 + T cells that constitutively express the Th2-like cytokines IL-4 and IL-10 [10,11]. Furthermore, HPV-11 E6 protein has been shown to cause the Th1/Th2 cytokine imbalance in papilloma and in peripheral blood mononuclear cells (PBMCs), manifested as upregulate the expression of IL-10 and IL-4 and reduce the expression of IFN-γ, IL-12 and IL-18 [12]. These altered T cell immune response, in combination with the skewed cytokine expression, may reveal the polarization of the adaptive immune system toward a Th2-like in RRP patients, which subsequently suppress Th1 cell-mediated clearance of the infection [10]. However, the functional pro le of HPV-speci c T-cell responses remain unclear, and these studies mainly focused on AORRP patients [13]. Indeed, JORRP usually runs a more aggressive clinical course, requiring a more thorough study for its antiviral immunity. In our previous studies, it has been revealed that children with JORRP exhibited a Th2-biased cytokine pro le [14]. Furthermore, we found that impaired T cell-dependent humoral immune response supported persistent HPV infection in JORRP [15]. These studies exposed the abnormal immune response of T cells to HPV in JORRP.
In this study, we performed a further assessment of circulating cytokines expression in JORRP, and explored the state of the antiviral immune response mediated by T cells. More important, Upon HPV6/11 antigens stimulation, functional changes in the circulating HPV-speci c T-cell responses were scrutinized, including T cell activation, proliferation, senescence, and apoptosis. We found that the HPV-speci c T cell immune response in JORRP children was impaired, which may be the major contributing factor to the recurrence of JORRP.

Subjects and sample collection
Demographic and clinical characteristics of JORRP patients and healthy controls are described in Table  1. Overall 46 JORRP patients from Beijing Children's hospital admitted at the period from January 2018 to December 2019 were enrolled. Ninety three age-and sex-matched healthy controls subjected to the routine physical examination between January 2018 and December 2019 at Beijing Children's Hospital were recruited. The study was approved by the Medical Ethics Committee of Beijing Children's Hospital, Capital Medical University (Grant No. 2019-k-43), which acted in compliance with ethical standards de ned by the Declaration of Helsinki. Written consents were signed by all the participants or their legal guardians. Peripheral blood samples from healthy controls and JORRP patients were collected in BD Vacutainer™ plastic blood collection tubes with EDTA K2 as anticoagulant. Plasma samples were collected by a centrifugation at 600 g for 5 min at room temperature and the supernatant were aliquoted and stored at -80 ℃.

Assessment of JORRP severity
To provide an overall assessment of disease severity, patients were categorized as having 'aggressive' or 'non-aggressive' disease according to Doyle et al. criteria as previously reported [15]. This binomial classi cation has been applied clinically by other researchers [16]. The characteristics of aggressive disease include 10 or more total procedures with three or more procedures within a 1-year period and/or spread of disease distal to the subglottis. By contrast, the characteristics of non-aggressive disease include less than 10 total procedures, less than three procedures within a 1-year period, and the absence of distal spread.

Plasma cytokine concentrations determined by a Luminex 200 system
Plasma concentrations of TNF-α, IFN-γ, IL-10 and IL-4 were determined using a human cytokine panel (MILLIPLEX MAP KIT) and read by a Luminex 200 system (Merck Millipore, Darmstadt, Germany). Luminex was performed as previously described [18]. All samples were measured in duplicate.

PBMCs isolation
Freshly isolated EDTA anticoagulated blood was diluted with PBS solution and layered carefully on Ficoll-Hypaque density gradients. After being centrifuged at 1000 g for 20 min interphase cell layer was carefully transferred into a 15 ml tube. Then the 15 ml tube was lled with 10 ml PBS, and cell pellet was collected after centrifugation at 500 g for 5 min.

Real-time PCR
To con rm the differential expression level of cytokines between patients and healthy controls, total RNA was extracted from PBMCs using the Direct-zol RNA Miniprep (ZYMO research, USA), and the rst strand cDNA was synthesized by a RevertAid First Strand cDNA Synthesis Kit (Thermo scienti c, USA). SYBR Green real-time PCR was performed with corresponding primers in a QuantStudio 6 ex real-time PCR system (Applied Biosystems). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was ampli ed as an internal standard for nal normalization. The primer sequence is listed in Table 2. The HPV6 and HPV11 recombinant antigens for in vitro cell culture were synthesized commercially (ProSpec, USA) as described in previous report [19]. Recombinant HPV6 antigen and HPV11 antigen are sequences of immunodominant antigens that are expressed in E. coli. Recombinant HPV6 antigen (Catalogue No. HPV 003) is a 55.6 kDa protein covering the full-length of HPV6 major capsid while recombinant HPV11 antigen (Catalogue No. HPV 004) is a 58.1 kDa protein covering the full-length of HPV11 major capsid (Table S1). Lyophilized antigens were dissolved in phosphate-buffered saline (PBS) at a concentration of 100 mg/ml (stock) and the aliquots were stored at -80°C for use. The titrated dose of 2.5 µg/ml of both antigens were used in culture.
In vitro stimulation of PBMCs and cell culture For each sample, the PBMCs were isolated, suspended in 10% FBS RPMI-1640 medium (Gibco, Invitrogen, Carlsbad, CA, USA) (5×10 5 cells/well) in 96-well plates (round-bottomed) at 37°C humidi ed cell incubator with 5% CO 2 . Recombinant HPV6 and HPV11 antigens (HPV6/11 antigens) were added to the culture for a 48-hr stimulation. All samples were further processed for cell surface staining with titrated uorochrome-labelled antibody.

Elispot assay
The HPV-speci c response was displayed with a human Elispot assay kits determining the number of responding cells releasing IFN-γ, IL-10, IL-4 or TNF-α after a 24-hr incubation with HPV6/11 recombinant antigens in 96-well round bottom plates (5×10 5 cells/well). The experiment was implemented according to the manufacturer's instruction [12]. In each assay, bovine serum albumin (BSA) was added instead of HPV antigen as a negative responder control, and cell mitogen phytohemagglutinin

Statistical analyses
The data were represented as the mean ± SD. Signi cant differences between JORRP patients and healthy controls were determined by the Student t-test or rank sum test according to the normal distribution test. Statistical analysis was performed with Prism (version 5.04) software. p < 0.05 was considered to be signi cantly different.

Results
Th1/Th2 cytokine imbalance correlated with disease severity in JORRP Cytokines expression pro le is one of the main components shaping the anti-infection immunity and mediating the pathogenesis of tissue damage. We were interested in knowing if pathogen and host interaction during HPV infection led to a skewed cytokine expression pro le in the JORRP peripheral system. With this aim, we rst measured the plasma concentration of the multiple cytokines from 46 JORRP patients and 93 healthy controls using a Luminex 200 system. The result indicated that the plasma concentration of TNF-α was signi cantly lower in JORRP patients than in healthy controls, while the IFN-γ, IL-10 and IL-4 were not evidently different (Fig. 1A). More importantly, we noted a signi cant negative correlation between the plasma concentration of TNF-α and the number of surgeries, the index for disease recurrence frequency. In contrast, IL-10 and IL-4 presented a positive correlation with the number of surgeries (Fig. 1B).
Secondly, in order to observe the transcription levels of these cytokines in circulating immune cells, we compared the mRNA expression in PBMCs from JORRP patients with those from healthy controls by realtime PCR. The results showed that JORRP patients had a down-regulated mRNA expressions of IFN-γ and TNF-α and an up-regulated mRNA expressions of IL-10 and IL-4 (Fig. 1C). Association analysis further revealed that IL-10 mRNA expression was signi cantly positively correlated with the number of surgeries, while IFN-γ, TNF-α and IL-4 presented no signi cant correlation (Fig. 1D).

JORRP patients had a positive memory T cell immune response to HPV
The view that the altered T cell immune response plays an important role in the progression of JORRP disease is now widely accepted [10]. We speculated that the changes of various cytokines in plasma and PBMC of JORRP children would closely relate to the behavior of T cells in peripheral blood. To test this, we measured the proportion of total T cells (CD3+) and T cell subsets, including CD4 + and CD8 + T cells in peripheral blood. The results showed that neither the percentage of total T cells in PBMC nor the ratio of any of T cell subpopulations changed in JORRP group compared with the healthy controls. (Fig. 2A-C).
Unexpectedly, we noted a signi cant reduction in the proportion of CD45RA expressing naïve T in JORRP patients compared to healthy controls. In contrast, the percentages of CD45RO expressing memory T cells were increased in JORRP patients ( Fig. 2D-F).
To further assess the responsiveness of T cells to HPV, we stimulated PBMCs isolated from JORRP patients or healthy controls with HPV6/11 antigens for 48h, and then analyzed the activation phenotype of T cells in PBMC by ow cytometry. Our result indicated that the active markers CD25 (Fig. 2G) and CD69 (Fig. 2H) were signi cantly increased in T cells from JORRP patients.

HPV induced the polarization of T cells to a Th2-like phenotype
The phenotype of the T cells was further analyzed for the expression of Th1/Th2 cytokines. Here we performed a cytokine-speci c elispot assay and then found that HPV6/11 antigens restimulation induced a higher number of IFN-γ, IL-10 and IL-4 secreting cells in PBMC from JORRP patients compared with that from the healthy controls, while the number of TNF-α secreting cells decreased (Fig. 3A-D). Of note, the number of IL-10 (892 ± 77) and IL-4 producing cells (832 ± 91) was signi cantly greater than the number of IFN-γ producing cells (240 ± 91) and TNF-α producing cells (723 ± 74) in JORRP patients (Fig. 3E). In other words, the PBMCs, primarily T cells, that respond to HPV antigen presented an Th2-like phenotypic polarization in JORRP children.

The proliferation, senescence and apoptosis of HPV mediated T-cell in JORRP
Now that we have known that there is a skewed phenotype of HPV antigen stimulated T cells in children with JORRP, we were eager to know if HPV could induce functional changes in T cells. First, we examined the effect of HPV on T cell proliferation. The isolated PBMC were labeled by CFSE and stimulated by HPV6/11 antigens in vitro for 3 days. The ow analysis results revealed that T cells from healthy controls generated more progeny through cell divisions than T cells from JORRP patients (Fig. 4A-B).
Next, ow cytometry was used to analyze the expression of cell surface marker CD57, a general marker of highly differentiated or senescent T cells [20]. We observed that the frequencies of CD57 in total T cells and memory T cells were increased in JORRP patients compared with healthy controls after HPV 6/11 antigens stimulation (Fig. 4C-D). As to apoptosis, based on our AV + assay by ow cytometry, the expression of AV in total T, naive T and memory T cells showed a remarkable increase in JORRP patients ( Fig. 4E-G), which was consistent with CD57-expressing T cells. Taken together, these analyses maybe suggest that HPV induce T cells to turn into a senescent phenotype and underwent apoptosis in a shorter time among JORRP patients.

Discussion
Activation and inhibition of the cellular immune responses are linked to cytokines expression in humans.
The dysregulated expression of cytokines in cellular immunity against HPV infection may facilitate the development of RRP. However, the current literature presented different cytokine expression pro les in RRP disease and lacked a systematic study for JORRP patients. In this study we perform a comprehensive evaluation of the expression of 4 representative T cell associated cytokines in 34 JORRP children (Fig. 1). Compared with that in healthy controls, the plasma level of TNF-α was found signi cantly reduced in JORRP patients and were negatively correlated with the number of surgeries. This agreed with ndings in AORRP which suggested a suppression of pro-in ammatory Th1 cytokines [12]. Interestingly, while IL-10 and IL-4 presented with comparable levels as those in control group, they exhibited a positive relationship with the number of surgeries. Surprisingly, different from other investigations [21], in our hand, we never tracked a signi cant decrease in plasma IFN-γ. This could be a re ection of undisturbed protein synthesis and secretion of IFN-γ in JORRP immune cells as our previous study. We have reported that upon a phorbol 12-myristate 13-acetate stimulation, the expression of intracellular protein level of IFN-γ in CD4 T cells was not altered in a noticeable way [15]. Nevertheless, in our study, the mRNA level of IFN-γ was shown dramatically decreased in PBMCs of patients. Indeed, there was a general reduction in the expression of pro-in ammatory cytokines at mRNA level in PBMCs from JORRP patients. Besides IFN-γ, the mRNA expression of TNF-α decreased in JORRP patients accompanied with an increase in the mRNA expression of anti-in ammatory cytokine IL-10 and IL-4, in which IL-10 presented a positive relationship with the number of surgeries. These results agreed with previous study in which mRNA of these cytokines was measured in PBMCs or papilloma of RRP patients [22,12]. In conclusion, the skewed cytokine expression in plasma and PBMC could be the synergic molecular modulations for the shift of the T cell polarity in response to HPV infection in JORRP patients.
The view that RRP patients establish "low level" tolerance [8] to the initial HPV-6 or -11 infection, and develop an exaggerated tolerogenic response to these viruses, rather than anti-HPV responses that effectively clear or contain them is now widely accepted. Cumulative evidence has suggested that T cells, as the most critical effector cells against HPV, are involved in the tolerance regulation of HPV [10]. Thus, understanding the mechanism(s) by which HPV-6 and − 11 polarizes the T cell immune response towards tolerance in RRP, as opposed to the development of cell-mediated immune clearance of these viruses, is critical in developing novel therapies that would prevent disease recurrence and/or reduce disease severity. In the present study, we did not observe the difference in the proportion of circulating total T cells (CD3+), CD4 + T and CD8 + T between JORRP children and healthy controls, which was in line with some of other studies ( Fig. 2A-C) [23]. Interestingly, we found that JORRP patients had an increased proportion of memory T cells and a decreased proportion of naïve T cells. This nding may attribute to the inhibitory cycle of immunocytes in RRP (Fig. 2D-F). The inhibition cycle model indicated that memory Th2-like T cells expressing IL-4, IL-10 and TGF-β alternatively activate macrophages to express the Th2-like chemokines CCL17 and CCL18, which polarize naïve CD4 + T cells to become memory Th2-like T cells and Tregs [10]. In light of that, we further analyzed the phenotypes of T cells stimulated by HPV6/11 antigens and found that compared with healthy controls, T cells in JORRP patients showed a higher HPV antigenic reactivity, manifested by up-regulated expression of activation markers CD25 and CD69 (Fig. 2G-H). In previous studies, CD4 + CD25bright CD127-Foxp3 + Treg cells expressing PD-1 and CD69 have been shown to be enriched in papillomas [24]. The functional Tregs would likely circulate to the periphery and suppress HPV-speci c T cell immune response [10]. Thus, upon HPV antigen restimulation, JORRP subject may present increased circulating CD25 bright and CD69 + Treg cells, which would trigger immune tolerance to HPV.
In addition, the phenotype of the T cells was further analyzed for the expression of Th1/Th2 cytokines. the cytokine-speci c elispot assay showed that the PBMCs, primarily T cells, that responding to HPV antigen presented an Th2-like phenotypic polarization in JORRP patients, with a dominant expression of IL-10 and IL-4 compared to IFN-γ and TNF-α, which would block effector Th1-like responses to HPV-6 and − 11 in patients (Fig. 3). James E et al. previously observed that HPV E2/E6-speci c CD4 + T-cell clones from the RRP subjects produced lower levels of IFN-γ and higher levels of IL-13 protein, also tended to exhibit reduced TNF-α secretion compared with healthy control subjects [25]. Taken together, all these evidences suggested that HPV prevents an effective anti-viral T-cell response in JORRP.
Although the phenotypic changes in T cell induced by HPV have been exposed, few studies have focused on the ultimate fate of T cells in chronic HPV infection. In our study, we found the expression of CD57 + and AV + T cells in JORRP patients responding to HPV6 /11 antigen was increased, while the proliferation of T cells was decreased (Fig. 4). Expression of CD57 usually represented the phenotype associated with replicating senescent T cells and is thought to be responsible for the inability of these T cells to proliferate [26, 27,28]. This proliferative defect was found in all CD57-expressing CD4 + and CD8 + T cells and NK cells and could not be overcome by the addition of exogenous IL-2 or IL-15 [29]. These T cells commonly were found in individuals with chronic immune activation [30,31], and they increased in frequency with age [32]. In Bonagura et al. 's study, the CD57 + CD4 + T-cell subsets were signi cantly elevated in papillomas of RRP susceptible population who lacked both KIR3DS1 and KIR2DS1 genes compared with patients who expressed one or both of these KIR genes [33]. It is precisely because of the evidence that restricted T-cell clones in papillomas circulate in the blood of RRP patients [10] that it is not di cult to understand that the increased CD57 + T cells were found in peripheral circulation of JORRP patients responding to HPV6 /11 antigen in our study ( Fig. 4C and 4D). These results may imply that chronic HPV infection induced T cells to turn into a senescent phenotype and underwent apoptosis.
Interestingly, we also found that memory T cells from JORRP have increased apoptosis and tended to exhibit increased expression of CD57 in response to HPV antigen restimulation (Fig. 4G) Supplementary Tables   Table S1 is not available with this version. Figure 1 Identi cation of the T cell associated cytokine levels in plasma and PBMCs of JORRP patients. (A) Circulating cytokine levels in plasma from healthy controls and patients with JORRP were analyzed by analysis between the number of surgeries and cytokine transcription levels in patients with JORRP. mRNA expression results of cytokines were represented by relative quanti cation (RQ). Spearman correlation coe cients (r and p-value) were shown. * p < 0.05, ** p < 0.01, and *** p < 0.001 compared with the healthy controls.