CD4⁺ and CD8⁺ T Cell Activation in Children with Hepatitis C

Abstract

Objectives

To assess if peripheral T cell populations in children with chronic hepatitis C virus (HCV) infection would show evidence of activation/exhaustion and an attenuated functional response.

Study design

Compared with adults, children with HCV infection have a higher rate of spontaneous viral clearance. In adults, chronic HCV has been linked to T cell exhaustion. Little is known of the immune status of children with HCV. Peripheral blood mononuclear cells were isolated from 16 children with HCV (6 males, 10 females; mean age 8.6 years, range 2-17), 16 age- and sex-matched control children without HCV infection, and 20 adults with chronic HCV. Multiparameter flow cytometry was performed to characterize T cell differences across the 3 groups.

Results

Controls and children with HCV had similar levels of CD4⁺, CD8⁺, and γδ⁺ T cells. Children with HCV demonstrated a decrease in naïve T cells compared with control children and increased activation/exhaustion marker expression on both CD8⁺ and CD4⁺ T cells. Transcription factor analysis suggested functional activation of T cells in children with HCV; however, only the CD4⁺ subset had enhanced cytokine production (interferon gamma and interleukin-2) compared with control children.

Conclusions

The HCV response in children is characterized by several changes in T cell phenotype. Many of these changes, such as increased T cell expression of programmed cell death-1, are similar to responses in adults. Of note, cytokine production by CD4⁺ helper T cells is increased in children with HCV compared with age- and sex-matched control children, which may influence long-term prognosis in children with HCV.

To date, pediatric hepatitis C virus (HCV) infection has received relatively little attention, but with an estimated 23000-46000 children infected, it will have a significant economic impact over the next 10 years.¹ Children usually acquire HCV infection through perinatal transmission, whereas adults commonly acquire hepatitis C through intravenous drug use or blood transfusion. When adults are acutely infected with HCV, they have an approximately 20% chance of clearing the virus.^2-4 The perinatal transmission rate of HCV is low; only 2%-6% of children born to mothers with HCV develop HCV.⁵ Of these children, 25%-40% will have spontaneous viral clearance, usually by age 2.⁶ Older children may subsequently clear the virus before adulthood, estimated at a rate of 6%-12%.^7,8 This higher rate of spontaneous clearance would suggest that children have improved mechanisms of viral clearance. However, the few children who are chronically infected may have a relatively weaker immune response, which keeps them chronically infected.

In adults, HCV clearance is mediated by both the adaptive and the innate immune systems.^9-11 The adaptive immune response to HCV is driven by T cells, as antibodies are not protective.¹² HCV clearance has been correlated with vigorous proliferation of CD4⁺ helper T cells with concurrent interleukin (IL)-2 and interferon (IFN)-γ production.¹³ The early, sustained development of CD4⁺ T cell response is critical for viral clearance, whereas only weak HCV-specific CD4⁺ T cell responses are observed in chronic infection.¹⁴ CD4 help during acute HCV infection is essential for spontaneous recovery¹⁵ because cytotoxic T lymphocyte priming in the presence of CD4 help is a critical factor in developing a protective response.¹³ Chronic HCV infection is characterized by an impaired HCV-specific cytotoxic (CD8⁺) T cell response that is unable to control replication.^14,16 Cytotoxic T cells play a major role in viral control during spontaneous infection resolution, however, these cells develop an exhausted status during chronic onset.¹⁷ In adults with HCV, both CD4⁺ and CD8⁺ T cells show increased expression of activation/exhaustion markers such as programmed cell death-1 (PD-1).¹⁸ PD-1 is involved in the downregulation of immune responses and binding of its ligand expressed on T cells results in inhibition and its expression is associated with viral persistence.^19,20 Several other T cell coinhibitory receptors have been recently associated with T cell exhaustion including T cell immunoreceptor with immunoglobulin and immunoreceptor tyrosine-based inhibition motif domains (TIGIT).²¹

Functional T cell exhaustion follows a defined pattern. T cells that undergo exhaustion first lose their capacity to produce IL-2, a cytokine that supports proliferation. IL-2 is predominantly produced by CD4⁺ T cells, whereas CD8⁺ T cells produce little IL-2 themselves and depend on CD4⁺ T cell help. This is followed by sequential loss of cytotoxicity and tumor necrosis factor a and IFN-γ production.²² T-box transcription factor TBX21 (T-bet) and eomesodermin (EOMES) are critical for IFN-γ production by T cells.²³ There is some evidence that HCV replication can suppress T-bet and IFN-γ signaling.²⁴ However, little is known about T-bet and EOMES in the setting of HCV, and these factors have not been studied in children.

Immune characteristics of children with HCV are relatively unexplored. Some studies suggest a role for innate immune cells including neutrophils,²⁵ natural killer,^26,27 and dendritic cells²⁷ in the pathogenesis of HCV in children. Other studies point to a role for T cells suggesting increased activation and a shift to an inflammatory (Th1) pattern of cytokine production in children with vertically transmitted HCV infection²⁸ and protection mediated by CD4⁺ T cell responses.²⁹ In another study, neonates exposed to HCV showed a relative suppression of immune activation, which was counterbalanced by an increased production of IFN-γ.³⁰

Based on the evidence from adults with chronic HCV infection and the limited data available for children, we hypothesized that peripheral T cell populations in children chronically infected with hepatitis C would show evidence of activation/exhaustion and an attenuated functional response compared with age- and sex-matched control children. We found the response in children with HCV to be characterized by several changes in T cell phenotype. Many of these changes, such as increased expression of PD-1, are similar to responses observed in adults. Of note, cytokine production by CD4⁺ helper T cells is increased in children with HCV compared with age- and sex-matched control children, which may influence long-term prognosis in children with HCV.

Methods

These studies were approved by the Institutional Review Board at the University of Colorado (Protocols 08-1350, 14-1809, and 06-0566). All adult subjects signed an informed consent before samples were taken. A parent or guardian of any child participant signed the informed consent on their behalf with assent obtained for children ages 8 years and older.

Children with HCV (n = 16) were recruited during liver clinic visits at Children's Hospital Colorado (Table; available at www.jpeds.com). Age- and sex-matched control children (n = 16) were recruited from the gastroenterology clinic or in the procedure suite prior to endoscopy (Table). The control children were being seen for a variety of general gastroenterology conditions but did not have underlying liver disease, inflammatory bowel disease, or did not require any immunosuppressive medications. Twenty adults who are HCV-positive served as an adult chronic HCV control group (Table). Blood was collected via standard blood draws into cell preparation tubes with sodium citrate (Becton, Dickinson and Company [BD], Franklin Lakes, New Jersey) and processed according to the manufacturer's protocol. Peripheral blood mononuclear cells (PBMCs) were viably frozen in 80% fetal bovine serum (BioWhittaker, Walkersville, Maryland), 10% dimethyl sulfoxide, and 10% Roswell Park Memorial Institute medium 1640 Media (Life Technologies, Grand Island, New York) and stored in liquid nitrogen for subsequent analyses.

Table.

Characteristics of children with HCV and control children and adults

Children with HCV
Age (y)	Sex	Race/ethnicity*	Genotype	AST†	ALT	HCV RNA‡	Biopsy (timing)§	Fibrosis stage
2	F	AA	1a	80	74	69 500	ND	NA
3.1	M	CA	1a	50	64	1 250 000	ND	NA
3.4	F	CA	1a	63	65	2 220 000	ND	NA
3.6	F	CA	1a	77	48	1 100 000	ND	NA
4.3	F	CA/H	1	50	43	214 000	ND	NA
4.8	F	CA/H	1	96	107	2 300 000	Same time	2/6 lshak
5.1	F	Unknown	1a	33	23	65 000	ND	NA
6.4	M	CA	1	147	124	1 045 523	ND	NA
6.9	F	CA/H	1a	94	86	914 000	4 y before and 2 after	2 of 4
9.3	M	Unknown	1a	52	60	High	10 mo before	1/4 trichrome
10.8	F	CA	1b	34	44	2 744 000	Same time	2/6 lshak
12.7	M	CA/H	1a	32	32	4 040 000	Same time	2/6 lshak
15.3	M	CA	1a	38	36	10 400	4 y before	1/6 lshak
15.9	M	CA	1	46	76	Detected	ND	NA
16.3	F	Unknown	1a	28	44	168 000	ND	NA
17.8	F	Unknown	1a	16	21	8 360 000	2 mo after	2/4 trichrome

Control children
Age	Sex	Race/ethnicity	Genotype	AST	ALT	HCV RNA	Underlying diagnosis
2.4	F	CA/H	NA	ND	ND	NA	Abdominal pain
3.2	M	CA	NA	59	27	NA	Eosinophilic esophagitis
3.8	F	CA	NA	ND	ND	NA	Foreign body
4.3	F	CA	NA	75	77	NA	Celiac antibodies
4.3	F	CA	NA	61	30	NA	Abdominal pain
5.3	F	CA	NA	53	24	NA	Abdominal pain
5.7	F	>1 race	NA	44	20	NA	Abdominal pain
6.7	M	AA	NA	50	26	NA	Eosinophilic esophagitis
6.7	F	CA	NA	40	27	NA	Celiac antibodies
9.3	M	CA	NA	64	23	NA	Abdominal pain
10.7	M	CA	NA	54	29	NA	Abdominal pain, diarrhea
12.5	F	CA	NA	ND	19	NA	Abdominal pain, celiac antibodies
14.8	M	CA	NA	29	15	NA	Abdominal pain
16.1	M	CA	NA	27	27	NA	Celiac antibodies
17.3	F	CA	NA	32	14	NA	Abdominal pain
17.9	F	CA/H	NA	19	18	NA	Abdominal pain, type 1 diabetes

Control adults with HCV
Age	Sex	Race/ethnicity	Genotype	AST	ALT	HCV RNA	Biopsy (timing)	Fibrosis stage¶
18	M	CA	1a	32	43	1 260 000	ND	NA
33	M	CA	1b	63	72	1930	ND	NA
24	M	CA	1a	36	63	10 400	ND	NA
50	F	AA	1a	26	22	ND	4.5 y before	Mild ¼
32	M	CA	1a	46	105	1 392 000	ND	NA
43	M	CA	1	75	117	986 000	ND	NA
55	F	CA	1b	85	96	1 720 000	ND	NA
43	F	CA	1a	46	52	2 200 000	4 y before	Periportal 2/4
56	M	Latino	1b	68	33	337 000	4.5 y before	Periportal 2/4
61	M	CA	1a	50	58	ND	1 y before	Fibrosis 2/4
37	M	CA	1a	27	22	1 590 000	8 mo before	Periportal 1-2/4
57	F	CA	1a	31	36	260 000	2 y before	Periportal 1/4
55	F	CA	1b	25	21	968 000	ND	NA
50	F	CA	1	35	42	ND	2.5	Fibrosis 2/4
55	F	CA	Unknown	42	42	ND	ND	NA
51	M	Latino	1b	82	61	<43	ND	NA
54	F	CA	1a	40	62	25 600	1 y before	Periportal 2-3/4
66	F	AA	1b	54	51	ND	ND	NA
58	F	CA	3	105	88	143 000	ND	NA
53	F	Latino	1b	80	52	2 860 000	ND	NA

AA, African American; ALT, alanine transaminase; AST, aspartate transaminase; CA, Caucasian; F, female; H, Hispanic; M, male; NA, not applicable; ND, not done.

^*Ethnicity non-Hispanic unless otherwise stated.

^†AST/ALT in U/L (normal <50).

^‡HCV RNA IU/mL.

^§Relative to time of blood draw.

^¶Trichrome stain.

Flow Cytometry

For cell surface staining, PBMCs were thawed and staining performed for 1 hour at 4°C. Antibodies used for surface staining included mouse anti-human CD3-V500 (Clone UCHT1), CD4-PerCP (Clone SK3), CD4-FITC (Clone RPA-T4), CD8-Allophycocyanin-H7 (Clone SK1), and rat anti-human CD197-FITC (CCR7) (Clone 3D12), purchased from BD. Additional antibodies included anti-human CD279- PerCP/Cy5.5 (PD-1) (Clone EH12.2H7; Biolegend, San Diego, California), and TIGIT-Allophycocyanin (Clone MBSA43; eBioscience, San Diego, California). Cells were fixed with 2% paraformaldehyde and acquired on a BD FACSCanto II Flow Cytometer. Analysis of flow cytometry data was performed using BD FACSDiva Software. Isotype controls were included for all samples. For intracellular staining of transcription factors, surface staining was performed as above. Intracellular staining was then performed using the Foxp3 staining buffer set (eBioscience) using the recommended protocol, except intracellular antibodies were incubated with the cells overnight at 4°C as this resulted in superior staining for T-bet and EOMES. Intracellular antibodies used were anti-T-bet R-phycoerythrin (Clone 4B10; Biolegend), and anti EOMES-FITC (Clone WD1928; eBio-science). Cells were acquired and analyzed as above, and intracellular isotype controls were included for all samples.

Functional Cytokine Assays

Thawed PBMCs 0.5 × 10⁶ in 200 uL were cultured in 96 well round-bottom plates. IL2 (25 U/mL) was added to all wells at the beginning of culture. Cells were cultured for 6 hours in the presence or absence of staphylococcal enterotoxin B (SEB, 1 ug/mL). Cells were then treated with Brefeldin A (1 ug/mL) and cultured for an additional 18 hours. At the end of the culture period, cells were harvested, surface stained with anti-CD3/CD4/CD8 to identify T cell subsets, followed by intracellular staining for IL-2 (R-phycoerythrin, Clone 5344.111; BD) and IFN-γ (V450, Clone B27; BD). The FIX and PERM Cell Permeabilization Kit (Invitrogen, Grand Island, New York) was used for intracellular cytokine staining. Cells were acquired and analyzed as above. Cells cultured in the absence of SEB served as background controls.

Statistical Analyses

Two-sided Student t test was used to compare differences between groups. Significance was defined as a P value of <.05. The JMP 6.0 (SAS Institute, Inc, Cary North Carolina) statistical software package was used.

Results

Sixteen children who are HCV-positive were included in this study. Characteristics and laboratory values are reported in the Table. Mean age was 8.6 ± 5.4 years. There were 10 females and 6 males. Mean aspartate transaminase was 58.5 ± 34 U/L, and mean alanine transaminase was 59.2 ± 29 U/L. Average HCV RNA level was 1 750 000 ± 2 250 000 U/mL. The 16-year-old female who developed HCV because of intravenous drug use was detected as having genotype 1 infection, but records from another hospital indicated that she previously had genotype 2 infection, was treated and cleared the virus, but 6 months after treatment became positive for HCV with genotype 1 infection. Four other pediatric patients had undergone treatment with ribavirin and IFN for HCV before entering the study: 3 were nonresponders, and 1 had a partial response. All patients had completed treatment at least 6 months prior to study enrollment. Thus, all the patients with HCV were viremic at the time of enrollment in this study. Some of the children with HCV had undergone liver biopsy around study collection time, and biopsy data where available is shown in the Table. Characteristics of the control children are shown in the Table. Mean age was 8.8 ± 5.4 years. Mean aspartate transaminase was 46.7 ± 16.6 U/L, and mean alanine transaminase was 26.9 ± 15.3 U/L. The adults with HCV in the control group are described in the Table.

Circulating Lymphocyte Populations in Children with HCV

Flow cytometric analysis was used to evaluate the levels of total T cells and subsets in the periphery of children with HCV compared with age- and sex-matched control children and adults with HCV. Levels of total T cells were similar in all study groups (mean 50.9%, 53.49%, and 50.26%, respectively, % of lymphocytes) as were levels of γδ-T cells (4.81%, 5.16%, and 2.96% respectively, % of total T cells). The proportion of CD4⁺ helper T cells was also similar (64.99%, 60.80%, and 65.75%, respectively, % of total T cells) in all study groups, as was the proportion of CD8⁺ cytotoxic T cells (27.51%, 25.54%, and 25.88%, respectively).

CD4⁺ Helper T Cell Phenotype in Children with HCV

Having established that the overall levels of and distribution of T cell subsets were similar; we next examined the naïve/memory phenotype of peripheral CD4⁺ T cells using the expression pattern of CD45RA and CCR7. Expression of CCR7 and the CD45RA isoform distinguishes naive T cells (CCR7⁺CD45RA⁺); and memory T cells, which are CD45RA⁻ and CCR7^+/−31 (Figure 1, A). Children with HCV had lower levels of naïve CD4⁺ T cells compared with control children, but levels remained higher than those detected in adults with HCV (Figure 1, B). Consistent with lower levels of naïve CD4⁺ T cells, expression of activation/exhaustion markers PD-1 and TIGIT (Figure 1, C and D) were increased on CD4⁺ T cells in children with HCV compared with control children, however, TIGIT level was lower than observed for adults with HCV. Upregulation of T-bet and EOMES in CD4⁺ T cells is associated with effector populations and act cooperatively to drive Th1 differentiation, IFN-γ expression and polyfunctionality.²³ We, therefore, looked at the expression of these transcription factors in CD4⁺ T cells in each of our study groups. Children with HCV infection had increased expression of both these transcription factors compared with age- and sex-matched control children (Figure 1, E and F).

Figure 1.

Flow cytometric analysis was used to characterize CD4⁺ helper T cells. A, Naïve cells are identified by the co-expression of CD45RA and CCR7. B, Naïve CD4⁺ T cell levels are decreased in children with HCV compared with age- and sex-matched control children but higher than those detected in adults with HCV. In children with HCV, CD4⁺ T cells, C and D, have a more activated phenotype, and E and F, express higher levels of transcription factors critical for functional maturation. FSC-A, forward scatter area; SSC-A, side scatter area.

CD8⁺ Cytotoxic T Cell Phenotype in Children with HCV

Having demonstrated several changes in the overall peripheral CD4⁺ T cell population in children with HCV, we next examined the phenotype of CD8⁺ T cells. As expected, levels of naïve (CCR7⁺CD45RA⁺) CD8⁺ T cells were decreased in children with HCV (46%) compared with control children (61%, P = .03) but were not significantly different from adults with HCV (36%). PD-1 and TIGIT expression patterns were similar to those on CD4⁺ T cells (Figure 2, A and B). Although EOMES expression mirrored that of CD4⁺ T cells, in contrast, no increase in T-bet was observed on CD8⁺ T cells (Figure 2, C and D).

Figure 2.

Flow cytometric analysis was used to characterize CD8⁺ cytotoxic T cells In children with HCV, CD8⁺ T cells, A and B, have a more activated phenotype and express higher levels of the C, transcription factor EOMES but D, not T-bet. *P < .05; **P < .01.

Cytokine Production

Our phenotype and transcription factor data suggested activation of peripheral T cells in children with HCV compared with age- and sex-matched control children, so we wanted to explore if these changes might have functional consequences. Therefore, we performed intracellular staining for IFN-γ and IL-2 after 24-hour stimulation with SEB. CD4⁺ T cells from children with HCV were more responsive, producing significantly more IFN-γ and IL-2. Of note, this group also had a higher proportion of CD4⁺ T cells that produced both cytokines (Figure 3). No difference in cytokine production was observed for CD8⁺ T cells.

Figure 3.

The functional capacity of pediatric T cells was assessed by intracellular cytokine staining for IFN-γ and IL-2 A, after 24-hour stimulation with SEB. CD4⁺ T cells from children with HCV were more responsive, producing significantly more B, IFN-γ and C, IL-2 and D, both cytokines. *P < .05.

Discussion

Little is known concerning the T cell status of children with chronic HCV infection. Children appear to experience an attenuated clinical course of infection as compared with adults.^32-34 In adults, the adaptive immune response to HCV is driven by T cells.^9,11,12 Highly functional CD4⁺ helper and CD8⁺ and cytotoxic T cell responses are critical for resolution of infection.^13-15 Conversely, chronic HCV infection is characterized by impaired or exhausted T cell responses demonstrating increased expression of activation/exhaustion markers such as PD-1.^14,16-18,20 In children, there is limited data that suggests that peripheral T cell responses are modulated by exposure to HCV in utero or chronic HCV infection.^28-30 However, only one of these studies included an age- and sex-matched uninfected control group for comparison with a group of children with chronic HCV.²⁸ Immune responses in children are thought to be decreased relative to adults and sex-related differences are prominent in early life^28,30,35,36; therefore, inclusion of an appropriate control group is of particular importance. Based on the evidence from adults with chronic HCV infection and the limited data available for children, we hypothesized that peripheral T cell populations in children chronically infected with HCV would show evidence of activation/exhaustion and an attenuated functional response compared with age- and sex-matched control children.

In the present study, we characterized peripheral T cell populations in 16 children with HCV and age- and sex-matched controls. We found that HCV infection in children induces several changes in T cell phenotype. Both CD4⁺ and CD8⁺ T cells were activated as evidenced by lower levels of naïve cells and increased expression of PD-1 and TIGIT compared with control children. Many of these changes, such as increased expression of PD-1, are similar to responses observed in adults with HCV. Of note, upregulation of TIGIT on T cells in children with HCV is attenuated compared with adults, which may contribute to the increased viral clearance rate observed in children. Alternatively, the somewhat decreased activation/exhaustion phenotype could be a reflection of the duration of infection. Because of the limited availability of samples from our child cohorts, we only examined expression of a limited number of activation/exhaustion antigens on T cells. Giovannetti et al²⁸ also found evidence of T cell activation with increased expression of Fas and HLA-DR on both CD4⁺ and CD8⁺ subsets. However, in contrast to our findings, they demonstrated a decrease in naïve CD8⁺ but not in CD4⁺ T cells. These differences could be related to sample size. Our data provides evidence of global peripheral T cell activation and suggests that comprehensive screening of co-stimulatory markers in the future may provide valuable information on the immune status of children with HCV.

We next looked for evidence of functional T cell exhaustion in children with HCV. Exhausted T cells first lose their capacity to produce IL-2, followed by sequential loss of cytotoxicity and tumor necrosis factor α and IFN-γ production.²² T-bet and EOMES are critical for polyfunctionality and IFN-γ production by T cells.²³ Therefore, we examined the expression of these transcription factors and performed intracellular cytokine staining for IL-2 and IFN-γ after stimulation with SEB. EOMES was increased in CD4⁺ and CD8⁺ T cells in children with HCV, but T-bet was increased only in the CD4⁺ T cell subset. In agreement with the transcription factor data, CD4⁺ T cells from children with HCV produced more IFN-y and IL-2 and had a higher proportion of CD4⁺ T cells producing both IL-2 and IFN-y, an indication of polyfunctionality. In their study, Giovannetti et al²⁸ demonstrated increased levels of CD4 and CD8 IFN-γ production, but saw no difference in the percentage of IL-2-producing T cells.²⁸ The use of phorbol myristate acetate and ionomycin for stimulation²⁸ as opposed to SEB may result in a response skewed toward IFN-γ production. It is important to note that these data relate to overall lymphocytes rather than HCV-specific responses. Because of the limited number of cells available and the difficulty in detecting HCV-specific responses in children with HCV^30,37 (our study focused on global T cell responses), further research is needed to elucidate the relevance of these findings to HCV-specific immunologic responses.

Age at time of infection is a critical factor in the development of chronicity and progression of HCV-related disease.³⁸ The rate of chronicity is lower in younger individuals.^39,40 Long-term follow-up studies in children with post-transfusion hepatitis indicate that only 55%-60% of children remain HCV RNA positive in adulthood.⁴¹ Although data is still lacking, it is thought that liver disease in children progresses slower than in adults. Of adults with chronic infection, 16% develop cirrhosis by 20 years after infection,⁴² and in a case series of 332 children, only 6 developed cirrhosis.⁴³ After infection, it takes time for an exhausted T cell profile to develop and the transition from acute to chronic infection can take several months. In children, given the resolution that we see in 2- and 3-year-olds, that conversion may take years. It is possible that children convert more slowly from an acute immune response to a chronic “exhausted” phenotype. However, we do not know if the 2- to 3-year-olds convert to a chronic phenotype and still clear the infection or if they remain in more of the acute immune reaction for several years and then clear the virus. Although our data do somewhat support a model of delayed T cell exhaustion in children, our study does not really address this question adequately. The duration of infection is shorter in our child cohort than in our adult cohort with HCV, which may also explain the “less exhausted” phenotype we see in children.

In summary, we found the response in children with HCV to be characterized by several changes in T cell phenotype. Many of these changes, such as increased T cell expression of PD-1, are similar to responses observed in adults. Of note, cytokine production by CD4⁺ helper T cells is increased compared with age- and sex-matched control children, which may influence long-term prognosis in children with HCV.

Glossary

BD

Becton, Dickinson and Company

EOMES

Eomesodermin

HCV

Hepatitis C virus

IFN

Interferon

IL

Interleukin

PBMC

Peripheral blood mononuclear cell

PD-1

Programmed cell death-1

SEB

Staphylococcal enterotoxin B

T-bet

T-box transcription factor TBX21

TIGIT

T cell immunoreceptor with immunoglobulin and immunoreceptor tyrosine-based inhibition motif domains