Role of TGF-β in anti-rhinovirus immune responses in asthmatic patients.

To the Editor:

The majority of viral infections of the airways are associated with asthma exacerbations in children. Two thirds of these viral infections are caused by rhinovirus, and hospital admissions for asthma correlate with the seasonal peak of rhinovirus infections.

TGF-β is a cytokine known to induce forkhead box P3+ (FoxP3) regulatory T (Treg) cells and retinoic acid–related orphan receptor (ROR) γt+ TH17 cells in combination with IL-2 or IL-6, respectively, but also to inhibit the differentiation of TH1 and TH2 cells.

Because TGF-β and rhinovirus infection both influence asthma exacerbation and TGF-β also induces rhinovirus replication, in this study we analyzed the effect of rhinovirus infection on TGF-β and the role of TGF-β on rhinovirus infection by analyzing asthmatic and nonasthmatic preschool children recruited in the European study Post-infectious Immune Reprogramming and Its Association with Persistence and Chronicity of Respiratory Allergic Diseases (PreDicta) and a murine model of asthma. The clinical data of the analyzed cohorts of children are reported in Table E1 and in the Methods section in this article's Online Repository at www.jacionline.org. In asthmatic children, in 66.6% of the cases, a viral infection was a triggering factor for development of the disease. Rhinovirus was the most common respiratory virus detected in the airways of these children (see Table E2 in this article's Online Repository at www.jacionline.org).

Table E1

Demographic and clinical data of the PreDicta cohorts WP1-UK-ER analyzed

Asthmatic patients								Control subjects
	Age (y)	Sex	Phenotype	Treatment	Asthma	Skin prick test	FEV1 (%)		Age (y)	Sex	FEV1 (%)
201	6	M	V	s + n (S+LTRA)	c	al, ca, g	126	208	6	M	77
202	6	M	U	st	p	al, b, g	80	211	6	F	121
203	5	F	U	st + ah	p	ca	101	214	5	M	112
204	6	M	A	st + n (S)	c	al, am, ca, g, h	128	215	4	M	—
205	5	M	U	st + n (S)	p	/	102	218	4	F	118
206	5	F	U	st	c	al	129	219	5	F	111
207	5	M	V	st + n (S)	p	g	143	220	5	F	96
209	4	F	A, V	st	p	g	115	221	3	M	81
210	6	F	V	n (S)	p	b, g	94	222	6	M	105
212	5	M	E, V	st + n (S)	p	—	98	226	4	F	109
213	4	M	E	st + n (S)	p	—	115	227	6	M	87
216	5	F	A, V	st + n (S)	u	ca, g, h	92	232	4	M	100
217	6	F	A, E, V	st	c	b, g, h	111	233	5	F	112
223	5	M	V	st + n (LTRA)	c	ca, f, g, h	99	234	5	F	119
224	4	F	V	st	c	—	135	235	4	M	116
225	4	M	V	st + n (S)	c	/	97	236	5	M	111
228	5	M	V	—	c	ca, g, h	84	237	4	M	109
229	4	M	V	n (S)	c	al, b, ca, g, h	86	240	4	F	92
230	5	M	V	—	c	al, am, b, ca, g, h	107	241	5	M	123
231	4	M	V	st + n (S)	c	b	71	244	5	F	107
238	4	M	V	st + n (S)	c	am, ca, g, h	86	245	4	M	121
239	5	F	E	n (ah)	c	/	98	246		M	109
242	5	M	A, E, V	st + n (ah)	c	/	81
243	5	F	V	st + n (S)	u	al, am, h	69
Average	4.9 ± 0.1						101.9 ± 4.0		4.7 ±0.18		105.7 ± 3.0

A, Allergen induced; ah, antihistamine; al, Alternaria species; am, ambrosia; b, birch; ca, cat; c, controlled; E, exercise-induced; F, female; f, Dermatophagoides farinae; g, grass pollen mix; h, house dust mite; LTRA, leukotriene antagonist; M, male; n, nonsteroid treatment; p, partially controlled; st+S, steroid treatment; U, unresolved; u, uncontrolled; V, virus-induced asthma.

Table E2

Viral colonization in nasal pharyngeal fluid (NPF) at baseline (B0)

Asthmatic patients	Rhinovirus	Control subjects	Rhinovirus
201	++	208	−
202	++	211	++
203	−	214	++
204	−	215	−
205	−	218	+++
206	++	219	+
207	−	220	−
209	−	221	++
210	CoVNL63++	222	−
212	Flu A++	226	+
213	++	227	+
216	−	232	+++
217	−	233	+, MPV+
223	+++	234	++, AdV++
224	++, PIV4+	235	++
225	++, PIV4+	236	−
228	−	240	+, PIV4+, HBoV+, FluA++
229	+++	241	++
230	+	245	−
231	HBoV++, CoVNL63+++	246	+
238	+++
239	+
242	+, HBoV+, RSVB++, HEV++, PIRV++, PIV4, AdV+
243	+++

In some cases other respiratory viruses were detected as follows: AdV, Adenovirus; CoV (NL63), human coronavirus (NL63); FluA/B, influenza virus A/B; HBoV, human bocavirus; HEV, enterovirus; MPV, human metapneumovirus; OC43/HK1, human coronavirus (OC43/HK1); PIV2/4, parainfluenza virus 1-4; RSVA/B, respiratory syncytial virus A/B; RV, rhinovirus.

+, Low copy number of rhinovirus; ++, intermediate copy number of rhinovirus; +++, high copy number of viral genomes detected in the sample; −, negative, no virus detected.

To investigate the role of TGF-β in rhinovirus-induced asthma in children, we analyzed PBMCs from preschool children with and without asthma, which were cultured for 48 hours after 1 hour of in vitro exposure to rhinovirus 1B (RV1B) and subjected them to gene array (Fig 1 , A, and see Table E3, Table E4, Table E5 in this article's Online Repository at www.jacionline.org). Because TGF-β induces Treg cells, we first investigated which genes related to tolerance were significantly regulated by rhinovirus in PBMCs from these children. Here we found that in asthmatic children rhinovirus upregulated immunosuppressive genes, such as cytotoxic T lymphocyte–associated protein 4 (CTLA4) and indoleamine 2,3-dioxygenase (IDO), programmed death ligand 1 (PD-L1; CD274), and interferon-induced transmembrane protein 2 (IFITM2; Fig 1, B and C). Consistent with the array data, we found that IDO1 was upregulated in PBMCs of asthmatic children cultured with rhinovirus compared with those of control children (see Fig E1, A, in this article's Online Repository at www.jacionline.org). This regulation was found to be independent from steroids because dexamethasone significantly downregulated IDO in PBMCs (see Fig E1, B).

Fig 1

PBMCs from asthmatic children exposed to RV1b in vitro upregulated IDO, PDL1, and LAP3. A, Experimental design for RNA arrays of PBMCs cultured in the presence or absence of rhinovirus (RV). B-E, Heat maps for asthmatic (Fig 1, B and C) and control (Fig 1, D and E) children and a differential expression analysis of the regulated genes are shown (asthma: n = 7, control: n = 5).

Table E3

Viral detection in nasal pharyngeal fluid (NPF) at the end point visit and 24 months (F4) after baseline (B0) in asthmatic and control children of the Erlangen cohort analyzed in gene arrays

Control subjects (n = 6)	F4	Asthmatic patients (n = 8)	F4
215	RV+, AdV+++	202	RV++
220	−	204	RV
221	−	205	RV+++
222	−	210	HEV
226	RV++	212	RV+
227	−	216	RV++
		217	RV++
		223	No sample

+, Low copy number; ++, intermediate copy number; +++, high copy number of viral genomes detected in the sample; −, negative, no virus detected; AdV, adenovirus; HEV, human enterovirus; RV, rhinovirus.

Table E4

Medications taken by the asthmatic children analyzed in gene arrays

Asthma F4	Rhinovirus in vitro	Medication
202	− and +	Steroid
204	− and +	Steroid + β2-agonists
205	− and +	Steroid + β2-agonist
212	− and +	Steroid + β2-agonists
217	− and +	Steroids
216	Only −	Steroids + β2-agonists
223	Only +	Steroid + LTRA

LTRA, Leukotriene antagonist.

Table E5

Meaning of the genes shown in the heat maps

CTLA4	Expressed by activated T cells. Other T cells can receive an inhibitory signal from CTLA4 to prevent an overreaction of the immune system.
IDO1	Tryptophan-converting enzyme that promotes T cell–mediated tolerance and antimicrobial effects. Several diseases are associated with increased expression.
CD274	Encodes an immune inhibitory receptor ligand. Can inhibit T-cell activation and cytokine production, which is essential for preventing autoimmunity (PD-L1).
IL1B	Encodes a protein that is a member of the IL-1 cytokine family. This cytokine is an important mediator of the inflammatory response and is involved in cell proliferation, differentiation, and apoptosis.
IFITM2	Encodes an interferon-induced antiviral protein that inhibits entry of viruses to the host cell cytoplasm, permitting endocytosis but preventing subsequent viral fusion and release of viral contents into the cytosol.
LDLR	Encodes a protein called a low-density lipoprotein receptor. This receptor binds to low-density lipoproteins, which are the primary carriers of cholesterol in the blood. Also, some rhinoviruses bind to the LDL receptor.
TLR8	Encodes a gene that is a member of the Toll-like receptor family. Takes part in pathogen recognition and activation of innate immunity.
MAD2L2	MAD2L2 is a component of the mitotic spindle assembly checkpoint that prevents the onset of anaphase until all chromosomes are properly aligned at the metaphase plate.
FAS	Encodes a protein belonging to the TNF receptor superfamily. It contains a death domain and has been shown to play a central role in the physiologic regulation of programmed cell death.
CDY2B	Encodes a protein containing a chromodomain and a histone acetyltransferase catalytic domain. Chromodomain proteins are components of heterochromatin-like complexes and can act as gene repressors.
TGIF2	Transcriptional repressor that can repress transcription by recruiting histone deacetylases to TGF-β–responsive genes.
RNASE1	Encodes a member of the pancreatic type of secretory ribonucleases, which cleave internal phosphodiester RNA bonds on the 3′ side of pyrimidine bases.
EVI5	An oncogene that can regulate the stability and accumulation of critical G1 cell-cycle factors, including Emi1 and cyclin A.
RHOG	Belongs to the family of G proteins and is involved in cellular signaling mechanisms, cytoskeletal reorganization, and subsequent morphologic changes in various cell types.
LY6E	Induced by IFN-α and is associated with drug resistance and tumor immune escape.
IL32	A proinflammatory cytokine that can induce cells of the immune system to secrete inflammatory cytokines, such as TNF-α and IL-6.
RNASE6	Belongs to the RNase A superfamily. Its expression is induced in neutrophils and monocytes on bacterial infection, suggesting a role in host defense.

Fig E1

MRNA expression of IDO in PBMCs. A,IDO mRNA expression in PBMCs from asthmatic (A) and nonasthmatic (CN) children with (+RV) or without (−RV) in vitro rhinovirus treatment. B, PBMCs from healthy volunteers were in vitro incubated with different concentrations of dexamethasone (Dex), 10−6 mol/L and 10−8 mol/L, and IDO expression was determined by using qPCR. *P ≤ .05 and **P ≤ .01.

Because TGF-β is secreted in a latent complex consisting of 3 proteins (TGF-β, the inhibitor latency-associated protein [LAP], and the ECM-binding protein LTBP), we also analyzed these and other TGF-β–inhibitory proteins. We noticed that TGF-β–inhibitory genes, such as TGIF2 and LAP3, were upregulated in rhinovirus-treated PBMCs from asthmatic children. Moreover, rhinovirus inhibited genes that cleave viruses, such as RNASE1, in PBMCs from children with asthma (Fig 1, B and C). By contrast, in control children rhinovirus did not significantly regulate these genes. In these children other factors were found to be significantly regulated by rhinovirus, such as lymphocyte antigen 6E (Fig 1, D and E), a protein involved in the TGF-β pathway. Moreover, we found that in PBMCs from control children, rhinovirus induced IL-32 (Fig 1, C and D). Expression of this protein is known to induce the production of IL-6 and TNF-α and might thereby modulate immune responses.

In subsequent experiments we analyzed in more detail the regulation of TGF-β in a larger group of children in the same cohort. Among PBMC supernatants, TGF-β protein was detected in high amounts in untreated cell-culture supernatants in both asthmatic and control children. However, after ex vivo challenge with rhinovirus, TGF-β protein expression was found to be significantly decreased (Fig 2 , A), although TGFB mRNA expression remained constant (Fig 2, B). Because rhinovirus infection suppressed TGF-β release, we assumed that rhinovirus facilitates TGF-β binding to the cell membrane, and for this reason, we could not detect it in the supernatants of rhinovirus-infected PBMCs.

Fig 2

Rhinovirus (RV) inhibits TGF-β release from PBMCs isolated from healthy and asthmatic children. A, TGF-β1 release from PBMCs of asthmatic and nonasthmatic children with or without in vitro rhinovirus infection analyzed by means of ELISA (n = 26-32 children per group, B0+F4). B-E, Relative mRNA expression of TGFB (Fig 2, B; n = 12-20), TGFBRII (Fig 2, C; n =3-6), FOXP3 (Fig 2, D; n = 19-31), and RORC (Fig 2, E; n = 19-31) in asthmatic and nonasthmatic children with or without in vitro rhinovirus infection (B0+F4) analyzed by means of real-time PCR. F-I, Correlation of RORC and FOXP3 mRNA expression in untreated and in vitro–infected PBMCs from asthmatic and nonasthmatic children. J and K, Relative TBX21 (Fig 2, J; n = 10-22) or IL6 (Fig 2, K; n = 10-27) mRNA expression from PBMCs in asthmatic and healthy children with or without in vitro rhinovirus treatment analyzed by using real-time PCR. The Student t test was used to calculate statistical significance. *P ≤ .05, **P ≤ .01, and ***P ≤ .001. Results are expressed as means ± SEMs.

To prove this concept of a viral immune escape mechanism, we analyzed the expression of TGFBRII in PBMCs in the presence or absence of in vitro rhinovirus infection. We found that PBMCs isolated from control and asthmatic children and infected with rhinovirus expressed increased levels of TGFBRII compared with the respective controls (Fig 2, C). This finding suggests that rhinovirus induced TGF-β receptor II expression, thus increasing TGF-β binding to the cell membrane and in this way explaining why we could not detect it in the cell supernatants.

To further analyze the influence of TGF-β signaling in molecules downstream of TGF-β, we then analyzed FOXP3 and RORC levels and found that PBMCs infected in vitro with rhinovirus express significantly more FOXP3 and RORC mRNA (Fig 2, D and E) in both control and asthmatic children.

When we analyzed the correlation of FOXP3 and RORC mRNA expression, we found a positive correlation of these 2 transcription factors in rhinovirus-challenged PBMCs in both groups of children (Fig 2, F-I). Taken together, rhinovirus infection induced FOXP3 and RORC.

We then asked whether T-box transcription factor (T-bet), a transcription factor known to regulate TH1/2, Treg, and TH17 cell development or activation, could be regulated by rhinovirus in PBMCs of children with asthma. Although we previously detected decreased TBX21 mRNA expression in asthmatic patients, here we found increased TBX21 mRNA levels in PBMCs isolated from asthmatic children after infection with rhinovirus compared with rhinovirus-infected PBMCs from control children (Fig 2, J). Thus TBX21 can be upregulated in asthmatic patients during active rhinovirus infection.

IL-6 is an inflammatory cytokine that, together with TGF-β, can induce the differentiation of TH17 cells. We found an upregulation of IL6 mRNA in control children after in vitro culture with rhinovirus. In contrast, asthmatic children showed a failure of such IL6 induction (Fig 2, K).

By analyzing naive and asthmatic mice, we found that in vitro treatment of lung cells with rhinovirus increased the proportions of TC1 cells, whereas adding TGF-β to the culture inhibited T-bet expression in CD4+ T cells, as well as IDO expression in total lung cells. The experimental set up, as well as the results, are described in detail in Figs E2 and E3 in this article's Online Repository at www.jacionline.org.

Fig E2

Rhinovirus induces TC1 cells in a murine model of asthma. A, Experimental design. i.n., Intranasal; i.p., intraperitoneal. B, CD8+IFN-γ+ cells were analyzed in total lung cells by using flow cytometry. A dot plot is shown for each group (n = 4-5 mice per group). *P ≤ .05, **P ≤ .01, and ***P ≤ .001.

Fig E3

TGF-β treatment of rhinovirus-infected cells reduces T-bet expression in CD4+ T cells in a murine model of asthma. A, Experimental design. i.n., Intranasal; i.p., intraperitoneal. B,Ido mRNA expression was detected in total lung cells from naive mice cultured with TGF-β after in vitro treatment with rhinovirus or untreated (n = 4). C, CD4+T-bet+ cells were analyzed in total lung cells from naive or OVA-treated mice by using flow cytometry. Exemplary dot plots are depicted for each group analyzed (n = 4-5 mice per group). *P ≤ .05 and **P ≤ .01.

In summary, these data suggest that in patients with acute rhinovirus infections, endogenous TGF-β is retained intracellularly in rhinovirus-infected cells, resulting in a T-bet–mediated immune response. At the moment, we do not know which cells are infected by rhinovirus in the PBMC population we examined; however, we assume that plasmacytoid dendritic cells are infected because of the induction of IDO after rhinovirus challenge ex vivo. However, rhinovirus infection also activates TGF-β present in the environment, as in patients with chronic asthma, to replicate and inhibit effective antiviral immune responses. Thus it is possible that children with acute asthma are able to induce an effective anti-rhinovirus immune response during acute exacerbation. By contrast, in patients with chronic asthma, TGF-β is increased in its active form and is released by structural cells. In this latter situation, when the rhinovirus infects plasmacytoid dendritic cells, this exogenous TGF-β inhibits TH1 and TC1 cells that carry the TGF-β receptor, resulting in TH1 cell depletion, and thus rhinovirus infection cannot be cleared. Although these data need further investigation, they open new avenues for our understanding of the role of rhinovirus-mediated asthma exacerbations in children.