A novel screening method detects herpesviral DNA in the idiopathic pulmonary fibrosis lung

Abstract

Background. Herpesviruses could contribute to the lung epithelial injury that initiates profibrotic responses in idiopathic pulmonary fibrosis (IPF).

Methods. We identified herpesviral DNA from IPF and control lung tissue using a multiplex PCR-and microarray-based method. Active herpesviral infection was detected by standard methods, and inflammatory cell subtypes were identified with specific antibodies. Patients that underwent lung transplantation were monitored for signs of herpesviral infection.

Results. A total of 11/12 IPF samples were positive for Epstein-Barr virus (EBV) and 10/12 for human herpesvirus 6B (HHV-6B) DNA. Control lung samples (n = 10) were negative for EBV DNA, whereas three samples were positive for HHV-6B. EBV-encoded RNA (EBER) was identified in nine IPF samples and localized mainly to lymphocytic aggregates. HHV-6B antigens were detected in mononuclear cells in IPF lung tissue. CD20+ B lymphocytic aggregates that were surrounded by CD3+ T cells were abundant in IPF lungs. CD23+ cells (activated B cells, EBV-transformed lymphoblasts, and dendritic cells) were observed in the aggregates. IPF patients had no signs of increased herpesviral activation after lung transplantation.

Conclusions. Inflammatory cells are the main source of herpesviral DNA in the human IPF lung. Diagnostic tools should be actively used to elucidate whether herpesviral infection affects the pathogenesis, progression, and/or exacerbation of IPF.

Key words::

• Herpesvirus
• interstitial fibrosis
• lymphocytes
• viral infection

Abbreviations
CD	=	cluster of differentiation
CMV	=	cytomegalovirus
EBV	=	Epstein-Barr virus
EBER	=	Epstein-Barr virus-encoded RNA
HHV	=	human herpesvirus
HSV	=	human herpes simplex virus
IPF	=	idiopathic pulmonary fibrosis
LMP-1	=	latent membrane protein 1
TGFβ1	=	transforming growth factor beta 1
UIP	=	usual interstitial pneumonia

Key messages

• Herpesviruses may contribute to lung epithelial injury in idiopathic pulmonary fibrosis.

• Inflammatory cells are the main source of Epstein-Barr virus and human herpesvirus 6B DNA in human idiopathic pulmonary fibrosis lung.

Introduction

Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive pulmonary disease with poor prognosis. Histopathologically, IPF lungs consist of usual interstitial pneumonia (UIP)-type lesions that are patchy parenchymal areas containing quiescent fibroblasts, fibroblastic foci, interstitial thickening with mononuclear inflammation, and meta-or hyperplastic epithelium. In advanced disease stages with the manifestation of honeycomb lung, the fibroblastic foci predominate, and extensive epithelial injury is observed. IPF patients should be evaluated early for lung transplantation, as no other effective treatment is available.

It is commonly believed that persistent lung epithelial injury initiates a profibrotic response leading to IPF/UIP (1). Infectious insult to the alveolar epithelium can be one contributing factor for epithelial injury in both IPF/UIP and IPF exacerbations (2). Evidence for an association between chronic viral infection and IPF/UIP has been gathered for different viruses including hepatitis C virus (3,4), cytomegalovirus (CMV) (5), and Epstein-Barr virus (EBV) (6–13). Recent data in mouse models of progressive lung fibrosis emphasize that herpesviruses can activate macrophage recruitment and transforming growth factor beta 1 (TGFβ1) expression that can be prevented by antiviral therapies (14–17). Furthermore, even latent murine gammaherpesvirus infection can augment experimental fibrosis (18).

This study was undertaken to see whether a newly developed multiplex polymerase chain reaction (PCR) and microarray-based assay for screening of eight different herpesviruses would be useful in the detection of viral DNA from human lung tissue samples. Our other aim was to examine whether the results could reveal an association between viral infection and IPF/UIP. As the majority of IPF/UIP samples included in the study were positive for EBV and human herpesvirus (HHV)-6B DNA, the samples were further analyzed using standard methods to evaluate infection: EBV in-situ hybridization and HHV-6 antigen detection. Immunostaining of inflammatory cell subtypes was performed to evaluate the possible host cells for latent herpesviral infection. Additionally, the proportion of patients that underwent lung transplantation was followed up after lung transplantation for evidence of reactivation of viruses that were detected in the explanted lung.

Materials and methods

Patient material

The lung biopsies were collected by an experienced pathologist from patients attending the Helsinki University Central Hospital. All included patients were informed of the study and gave written consent. The ethical committee of the Helsinki University Central Hospital approved the study, and it has been registered at www.hus.fi/clinicaltrials. Individual patient characteristics are shown in Table I. Control specimens consisted of healthy pulmonary tissue from explanted, unused organ donor tissue, or lobectomy due to benign and localized malignant tumors. One control patient had mild smoking-related emphysema. For the detection of viral antigens from peripheral blood in lung transplant patients, a data set of patients including the patients in this study was used (19).

Table I. Patient demographics and results for the viral screen.

Patient	Dg/PAD	Age	Sex	Smoking/Pack-your	Immunosuppression, mg	FVC%
Patient 1	UIP, TX	60	M	Never	Pred 5	41
Patient 2	UIP, TX	62	M	Never	Pred 10	n/a
Patient 3	UIP, TX	62	M	Ex/20	Pred 5	33
Patient 4	UIP, moderate	68	M	Ex/40	Medrol 4, Trexan 15/w	79
Patient 5	UIP, spotty	62	M	Never	None	83
Patient 6	UIP, epit. metaplasia, inflammation, TX	62	M	Ex/10	Pred 5	40
Patient 7	UIP, moderate	78	M	Never	None	46
Patient 9	UIP, TX	53	M	Ex/30	Pred 10	57.8
Patient 21	UIP, TX	59	M	Ex/30	Pred 20	38
Patient 22	UIP, TX	38	F	No	Medrol 40, Cymevene iv	31
Patient 23	UIP, TX	36	M	No	CyA 125, Pred 1040	45
Patient 24	UIP, TX	48	M	No	Pred 10	28
Patient 10	Emphysema	46	F	Smoker >40	None	89
Patient 11	Lymph node resection, normal lung	67	M	Ex/20	None	n/a
Patient 12	Localized adenocarcinoma	69	M	Ex/10	None	90
Patient 13	Hamartoma, normal lung	59	M	Ex/20	None	108
Patient 14	Hamartoma, normal lung	58	M	Ex/20	None	128
Patient 15	Histologically normal lung from a multiple organ donor	Unknown	Unknown	Unknown	Unknown
Patient 16	Normal lung from local carcinoma resecate	64	M	Ex/30	None	76
Patient 17	Multiple organ donor, normal lung	Unknown	Unknown	Unknown	Unknown
Patient 18	Emphysema	72	F	Never	None	95
Patient 19	Hamartoma, normal lung	46	M	Never	None	71
Patient 20	Melanoma malignum, normal lung	70	M	Never	None	88

Pred = prednisolone; CyA = cyclosporin A; epit = epithelial; Ex = ex-smoker; TX = lung transplant.

The diagnosis of IPF was made according to the ATS/ERS guidelines (20,21). Lung biopsies were taken either for diagnostic purposes or due to lung transplantation. The explanted lungs, open lung biopsies, or video-assisted thoracoscopic biopsies were kept at +4°C during transportation from the operating theater to the laboratory (time ranging from 1 to 3 hours). As shown in Table I, the transplant patients exhibited more advanced disease stages (lower forced vital capacity [FVC]). After all necessary diagnostic samples were collected, macroscopically representative samples were cut from excess lung tissue, snap-frozen immediately with liquid nitrogen, and stored at −70°C.

Oligonucleotide microarray for detection of eight different herpesviruses

Nucleic acids were extracted from deep-frozen lung tissue samples using Qiagen's Tissue kit (Qiagen Nordic, Helsinki, Finland) according to manufacturer's instructions. Prove-it™ Herpes-kit (version for Research Use Only, Mobidiag Ltd, Helsinki, Finland) that includes a separate multiplex-PCR and a microarray for detection of labeled PCR products (originally modified from Jääskeläinen et al. 2006 (22) and Jääskeläinen et al. 2008 (23)) was used to detect herpesvirus DNA. A multiplex-PCR and microarray (Prove-it™ Tube Array system, Mobidiag Ltd, Helsinki, Finland) were carried out according to the manufacturer's instructions. The assay allows detection of CMV, EBV, human herpes simplex viruses type 1 and type 2 (HSV-1, -2), Varicella-zoster virus, and human herpesvirus 6A, 6B, and 7 (HHV-6A, -6B, and -7, respectively) DNA. Results were analyzed with Prove-it™ TubeArray system. All the samples were run as duplicates, and the microarray reactions for the positive samples were carried out at least twice. The results were consistently reproducible.

Epstein-Barr virus detection

In-situ hybridization for EBV-encoded RNA (EBER) was conducted on formalin-fixed paraffin sections using a fluorescein isothiocyanate (FITC)-labeled oligonucleotide probe supplied by Ventana (Tucson, Arizona) on an automated immunostainer (Ventana Benchmark LT). Visualization was achieved using the ISH iView system with alkaline phosphatase and NBT/BCIP (nitro-blue tetrazolium chloride/ 5-bromo-4-chloro-3′-indolyphosphate p-toluidine salt) substrate, with Fast red as contrast.

CMV and HHV-6B antigen detection

CMV pp65 and HHV-6B antigens were demonstrated in the frozen sections of tissue specimens as described (19,24). For CMV pp65, the primary antibody dilution was 1:5 (Biotest pp65 Ab), and for HHV-6B 1:200 (Chemicon MAB 8535). Briefly, the freshly cut frozen sections were fixed with ice-cold acetone at −20°C for 10 minutes. The primary, secondary, and third antibodies were incubated each in room temperature (RT) for 30 minutes, and the Horseradish-peroxidase (HRP)-conjugated third antibody was visualized using AEC (3-aminoethylcarbazole). Demonstration of HHV-6 antigens in tissue specimens was performed as described. From peripheral blood, a similar staining protocol for HHV-6B and CMV antigen detection was used to cytocentrifuged peripheral blood leukocytes (19,25).

Immunohistochemistry

Immunohistochemistry for leukocyte CD antigens was carried out from paraffin sections using the Dako Autostainer System with the EnVision™ Detection Kit (K5007; Dako, Glostrup, Denmark), and 3,3′-diaminobenzidine (DAB) as the chromogen. Sections were deparaffinized in xylene, rehydrated in graded alcohol series, and pretreated with Tris-EDTA buffer (pH 9) at 98°C for 24 min. The slides were incubated with the following antibodies for 30 min: CD20 (1:100, Novocastra, Novocastra, Newcastle Upon Tyne, UK), CD23 (1:1000, DAKO), CD3 (1:10, Novocastra). The number of CD20-positive aggregates was calculated by counting their total numbers from each individual section.

Results

EBV and HHV-6B DNA is abundant in lung tissue of IPF patients

The results from the herpes DNA screening method for individual patients are shown in Table II. All except one IPF/UIP patient lung (n = 12) were positive for EBV DNA, whereas 10/12 IPF/UIP lungs were positive for HHV-6B DNA. In addition, three IPF/UIP samples were positive for HHV-7 DNA, two samples for CMV DNA, and one sample for both HSV-1 and HSV-2 DNA. Out of the control specimens (n = 10), none were positive for EBV DNA, whereas three samples were positive for HHV-6B DNA and two for HHV-7 DNA.

Table II. Comparison of the herpes DNA screening method with the standard methods of detecting EBV, HHV-6B, and CMV. EBV in-situ hybridization (EBER) positivity on paraffin sections was calculated individually from different compartments of the pulmonary tissue. Frozen sections were used for HHV-6B and CMV detection. Weak positivity is indicated in parentheses.

		EBER
Patient number	Herpesviral DNA screen	Mo	AEC	Ly	HHV-6B +	CMV pp65 +
1 UIP	EBV, HSV-1, HSV-2, HHV-6B, HHV-7	−	+ + +	+ +	+ + +	(+)
2 UIP	EBV, HHV-6B, HHV-7	+	+	+ +	n/a	n/a
3 UIP	EBV	−	+	+	n/a	n/a
4 UIP	EBV	+ + +	+	−	n/a	n/a
5 UIP	EBV, HHV-6B	−	+	−	n/a	n/a
6 UIP	EBV, HHV-6B	+	−	−	n/a	n/a
7 UIP	EBV, HHV-6B	+ + +	−	−	n/a	n/a
9 UIP	EBV, HHV-6B	−	−	−	+	(+)
21 UIP	HHV-6B, CMV	−	−	+	+	+
22 UIP	EBV, CMV, HHV-6B, HHV-7	−	−	−	+	+
23 UIP	EBV, HHV-6B	−	−	−	n/a	n/a
24 UIP	EBV, HHV-6B	−	+	−	n/a	n/a
10 Ctrl	HHV-7	+	−	−	n/a	n/a
11 Ctrl	HHV-6B, HHV-7	−	−	(+)	n/a	n/a
12 Ctrl	HHV-6B	−	−	(+)	n/a	n/a
13 Ctrl	NEG	−	−	−	n/a	n/a
14 Ctrl	NEG	−	(+)	(+)	n/a	n/a
15 Ctrl	NEG	−	−	−	n/a	n/a
16 Ctrl	NEG	−	−	−	n/a	n/a
17 Ctrl	NEG	−	−	−	n/a	n/a
18 Ctrl	NEG	−	−	−	n/a	n/a
19 Ctrl	NEG	−	−	−	n/a	n/a
20 Ctrl	(HHV-6B)	−	−	−	n/a	n/a

Mo = macrophages; AEC = alveolar epithelial cells; Ly = lymphocytes; n/a = not available.

Inflammatory cell aggregates are a possible source of EBV DNA in the IPF/UIP lung

As the majority of the IPF/UIP biopsies in this study showed positivity for EBV DNA, we performed EBER in-situ hybridizations from histological cross-sections to determine whether viral EBV-encoded RNA localizes to specific structures in the lung. EBER-positive cells were observed in five UIP biopsies (Table II). The positive cells were predominantly observed in parenchymal inflammatory cell aggregates (Figure 1A), but some specimens showed faint positivity in alveolar macrophages (Figure 1B) and in hyperplastic epithelial cells, that are often detected above the fibroblastic foci (Figure 1C).

Figure 1. Representative photomicrographs of Epstein-Barr virus-encoded RNA (EBER) in-situ hybridization of lung biopsies from patients with IPF. The positive cells were seen predominantly in the parenchymal lymphocytic aggregates (A, arrow points at a positive mononuclear inflammatory cell, 300× magnification), but some specimens showed faint positivity in alveolar macrophages (B, arrow points at intracellular faint positivity, 1000× magnification). C shows a faintly positive activated alveolar epithelial cell nucleus (1000×) above a fibroblastic focus. D shows a representative photomicrograph (300×) of HHV-6 immunoreactivity in a frozen section from a patient with IPF/UIP. HHV-6B antigens were detected in parenchymal mononuclear inflammatory cells, morphologically appearing as macrophages or lymphocytes.

Parenchymal mononuclear cells show positivity to HHV-6 antigen in the IPF/UIP lung

HHV-6B DNA was found in the majority (11/12) of IPF/UIP patient samples. To reveal signs of active HHV-6 infection and to localize possible viral particles in the lung, HHV-6B antigen detection was performed. Four patient samples positive for HHV-6B DNA were applicable for the studies, as frozen sections from only these patients had been collected from the explanted lung at transplantation. All of the samples were positive for HHV-6 antigens that were detected in parenchymal inflammatory cells morphologically appearing as macrophages and lymphocytes (Figure 1D). Positivity was dispersed evenly in the parenchyma, even at fibroblastic foci, and did not appear to be concentrated to inflammatory cell aggregates. The two samples positive for CMV DNA were also positive for CMV pp65 antigens (Table II).

Inflammatory cell clusters in the IPF/UIP lung are abundant in B cells

As EBV RNA and HHV-6 antigens seemed to be localized to parenchymal inflammatory cell aggregates and solitary mononuclear cells, we identified the inflammatory cell subpopulations from the UIP lung cross-sections by using standard markers for T cells, B cells, and dendritic cells. We used antibodies against CD20 (B cell), CD23 (activated B cells, dendritic cells, and EBV-transformed lymphoblasts), and CD3 (T cells).

All UIP biopsies had numerous CD20-positive B lymphocytic aggregates (Figure 2A, Table III) whereas only a few isolated peribronchiolar or subpleural B lymphocyte aggregates were found in the control specimens (Figure 2B). Clonality in the CD20-positive B cell clusters was ruled out using light chain κ and λ in-situ hybridization (both subchains were present in all UIP biopsies, data not shown). CD23, often detected in EBV-transformed activated lymphoblasts, was occasionally positive in the aggregates, but the CD23-positive cells had the morphology of dendritic cells (Figure 2C). The CD20-positive B cells were small and did not have the morphological characteristics of lymphoblasts. T cells surrounded the B lymphocytic aggregates and were also found dispersed in the lung parenchyma (Figure 2D).

Figure 2. Photomicrographs of lung biopsies from a patient with IPF (A), and control lung (B) stained with the antibodies against the B-lymphocyte marker CD20. Photomicrographs of lymphoid aggregates in IPF lung stained with antibodies against CD23 (activated B cells, dendritic cells and EBV-transformed lymphoblasts) (C), and CD3 (T cells) (D). Original magnification 300×.

Table III. Immunohistochemical analysis of paraffin sections using antigens for mononuclear cells (CD20 for B cells, CD23 for activated B cells and dendritic cells, CD3 for T cells).

Patient number	CD20+ clusters	CD23+ clusters	CD3+ clusters
1 UIP	32	4	25
2 UIP	20	1	10
3 UIP	21	0	21
4 UIP	12	(1)	12
5 UIP	8	3	8
6 UIP	25	5	15
7 UIP	12	0	13
9 UIP	3	0	0
21 UIP	35	1	>35
22 UIP	5	(1)	>5
23 UIP	(1)	(1)	>1
24 UIP	49	5	44
10 Ctrl	4	1	2
11 Ctrl	1	0	3
12 Ctrl	0	0	4
13 Ctrl	3	2	1
14 Ctrl	0	0	1
15 Ctrl	3	0	2
16 Ctrl	0	0	0
17 Ctrl	0	0	0
18 Ctrl	2	0	2
19 Ctrl	0	(1)	3
20 Ctrl	0	(1)	(1)

Lung transplant patient monitoring

Some patients underwent lung transplantation after the screening of viral DNA from the explanted lung. We hypothesized that if these patients had a clinically relevant or systemic latent virus infection at the time of transplantation, the infection would be activated after the administration of forceful immunosuppression. Altogether seven lung biopsies were taken from IPF patients from the explanted lung during lung transplantation. These patients were carefully monitored according to the local clinical protocol after transplantation for the activation of CMV or EBV that are known to be common infectious complications due to immunosuppression. In addition, the patients were followed up for the activation of HHV-6B and HHV-7 by detecting the virus-specific antigens in peripheral blood mononuclear cells. The technical details of the HHV-6B and HHV-7 antigenemia tests have been described previously (25). Most patients had signs of CMV infection after transplantation, but only one patient developed EBV infection. HHV-6-positive peripheral blood cells were detected from 5/7 patients, and the first positive cells in all patients were detected during the valacyclovir/gancyclovir prophylaxis. However, there were no differences in the frequency or incidence of HHV-6 between patients that underwent transplantation due to pulmonary fibrosis or other diagnosis. Only one UIP lung biopsy specimen exhibited positivity for HSV-1 and HSV-2 DNA. The HSV-1 and HSV-2 viruses were followed from bronchoalveolar lavage taken at routine controls, and no pulmonary HSV infections were found after transplantation.

Discussion

Although the etiology of IPF still remains unclear, one line of research suggests that viral infection is a possible etiologic and/or risk factor for IPF/UIP. Our group has previously reported that the susceptibility gene for Finnish familial IPF, ELMOD2, is essential for pathways in antiviral response (26). IPF is a relatively rare condition with the prevalence of 16–18/100,000 individuals in Finland (27). In contrast, most people (>95%) are infected by, e.g., EBV and HHV-6 at some point of their life. Hence, it is obvious that these infections do not normally lead to fibrotic events but may act as a trigger leading to an abnormal antiviral response in patients otherwise prone to fibrosis.

Here we used a novel multiplex-PCR and microarray-based assay for the detection and identification of eight herpesviruses from deep-frozen lung tissue samples. Originally, the assay was designed for the detection and identification of herpesvirus DNA from several kinds of clinical samples, e.g. cerebrospinal fluid, whole blood, plasma, and serum samples (22,23). Our strategy was to use the modified assay for screening of herpesvirus DNA-positive samples and continue by searching for signs of active infection. The results suggest that the assay is applicable for tissue samples as starting material. All IPF/UIP samples were positive for at least one herpesvirus, and the majority of them were positive for EBV DNA. A novel finding was that HHV-6B DNA was abundant in IPF/UIP lung. EBER in-situ hybridization and immunostaining of HHV-6B and inflammatory cell subtypes revealed that mononuclear inflammatory cells are a potential source for viral DNA. As most control specimens were negative for viral DNA, it is unlikely that positivity is due to peripheral blood contamination. Most IPF patients received corticosteroids that can be one contributing factor for viral activation.

In recent reports (6–13), 44%–96% of the analyzed IPF/UIP lung specimens have been positive for EBV DNA or active/latent proteins as typically detected by PCR or immunohistochemistry, respectively. Our results are in line with the previous studies. EBV DNA was detected in 11/12 UIP biopsies, and EBER-positive inflammatory cells were found in 9/12 samples. The detection limit of the herpes DNA screening assay is approximately 10–30 viral DNA copies in cerebrospinal fluid samples (22,23), and, according to the manufacturer, the sensitivity and specificity are 90% and 98%, respectively, in comparison to the PCR-based reference methods. Although DNA positivity and active infection are two different entities, comparison between the novel herpes DNA screening method and EBER in-situ hybridization is potentially interesting from a diagnostic perspective. The similar results obtained with these two methods suggest that the presence of viral DNA in the IPF/UIP lung can be indicative of viral infection.

In contrast to our findings that EBV resides mainly in pulmonary mononuclear cell aggregates, others have shown latent membrane protein 1 (LMP-1) in the alveolar epithelial cells of the IPF/UIP lung (9). Epithelial LMP-1 positivity has been associated with a poor prognosis in 29 IPF patients (12). Although EBV can replicate in the lower respiratory tract (8,10), it is still unclear how EBV enters the epithelial cells (28).

IPF is often referred to as being a non-inflammatory disease, possibly because it has little or no response to immunosuppressive treatment. As LMP-1-positive cells have mainly been found in epithelial cells, the localization of EBER positivity in the parenchymal lymphocytic aggregates and the abundance of B cells in IPF/UIP lungs were surprising. B cell aggregates in IPF/UIP lungs, however, are not a new finding; Wallace et al. detected aggregates of B lymphocytes in 37/38 IPF patients (29), and the original observation on lymphocyte subpopulations in IPF was made already in 1985 (30). More recently, lymphoid aggregates were suggested to provide the first evidence of lymphoid neogenesis in IPF/UIP (31). Similar aggregates referred to as lymphoid follicles are observed in advanced stages of chronic obstructive pulmonary disease (COPD) (32), indicating a common pattern of reaction to injury in the lung parenchyma. Others have shown that the number of CD3+ cells in IPF/UIP correlates to patient survival (33). It is likely that mononuclear cells are a major source of herpesviral DNA in the IPF/UIP lung, and changes in the pulmonary parenchymal microenvironment may lead to activation of these viruses and further enhance disease progression. Although a recent microarray study involving acute exacerbations of IPF showed no evidence of viral etiology of exacerbations (34), the activation of latent viruses is still a plausible explanation of acute IPF exacerbation and should be further explored with specific diagnostic tests for HHV-6 and EBV during exacerbation.

This is the first study to show a relationship between IPF/UIP and HHV-6. HHV-6 belongs to the subfamily of betaherpesviruses that includes CMV and HHV-7. HHV-6 has two closely related variants, A and B, of which the B variant is more common. Their association with lung diseases is poorly known, but HHV-6B is considered to be a pathogen in immunosuppressed transplant recipients. HHV-6B antigens have been detected in a case of lymphoid interstitial pneumonia (35). HHV-6 has immunomodulatory effects, and reactivation is often found in conjunction with other herpesviruses, such as CMV. Hence HHV-6 may be a co-factor promoting the reactivation of other herpesviruses. The presence of HHV-6 DNA in tissue samples from IPF patients has been investigated in one previous study, where the authors did not distinguish between the two HHV-6 variants and were unable to detect HHV-6 DNA from lung tissue (7). In this study, nearly all IPF patients had HHV-6B DNA in the lung. More importantly, we were able to show HHV-6B viral antigens indicating activation of the latent infection.

In conclusion, we show here that inflammatory cells are an abundant source of herpesviral DNA in the IPF/UIP lung, and in suitable conditions these viruses do activate. It is still unknown whether herpesviral activation can lead to disease progression or exacerbation. However, the current evidence certainly encourages clinicians to look actively for markers for systemic viral infection. HHV-6 antigen detection and HHV-6 and EBV RNA quantitation from peripheral blood are readily available as routine laboratory tests and might provide additional evidence for or against therapeutic antiviral attempts.

Acknowledgements

We would like to thank Eriika Wasenius and Tiina Marjomaa for their skilful laboratory work.

Declaration of interest: This study has been supported by the Academy of Finland, Sigrid Jusélius Foundation, The Finnish Anti-Tuberculosis Association Foundation, Helsinki University Central Hospital (HUS-EVO), Jalmari and Rauha Ahokas Foundation, and Finska Läkaresällskapet. H. Piiparinen and A. Jääskeläinen are both former employees of Mobidiag Ltd. The other authors declare no conflicts of interest.