Journal of Clinical Immunology

, Volume 31, Issue 2, pp 265–271

Novel STAT1 Alleles in a Patient with Impaired Resistance to Mycobacteria

Authors

  • Ines Ackerl Kristensen
    • Department of PediatricsCopenhagen University Hospital
  • Jens Erik Veirum
    • Department of PediatricsAarhus University Hospital
  • Bjarne Kuno Møller
    • Department of Clinical ImmunologyAarhus University Hospital
    • Department of Clinical ImmunologyAarhus University Hospital
Article

DOI: 10.1007/s10875-010-9480-8

Cite this article as:
Kristensen, I.A., Veirum, J.E., Møller, B.K. et al. J Clin Immunol (2011) 31: 265. doi:10.1007/s10875-010-9480-8

Abstract

Partial defects in interferon (IFN)-γ signaling lead to susceptibility to infections with nontuberculous mycobacteria. The receptors for IFN-α and IFN-γ activate components of the Janus kinase-signal transducer and activator of transcription (STAT) signaling pathway. Some defects in STAT1 mainly affect IFN-γ signaling, thus resulting in mendelian susceptibility to mycobacterial disease (MSMD). MSMD is a severe disease but patients show a favorable response to anti-mycobacterial chemotherapy. Other defects in STAT1 affect both IFN-α and IFN-γ signaling resulting in mycobacterial and lethal viral disease. We report here a patient with two novel STAT1 alleles, which in combination results in a recessive trait with partial STAT1 deficiency and mycobacterial disease. Cells from the patient did respond to mycobacterial antigen, but both the expression of STAT1 and phosphorylation of STAT1 in response to IFN-γ treatment were reduced. This is the first report of a mutation in the N-terminal part of STAT1 involved in causing mycobacterial disease.

Keywords

Humanimmunodeficiency diseasestranscription factorssignal transductionnontuberculous mycobacteriaMSMDSTAT1

Abbreviations

IFN-γ

Interferon-γ

MSMD

Mycobacterial disease

JAK

Janus kinase

STAT1

Signal transducer and activator of transcription-1

IFNGR

IFN-γ receptor

GAF

Gamma-activated factor

GAS

Gamma-activated sequences

ISGF3

IFN-stimulated genes factor 3

ISRE

IFN-α sequence response element

PHA

Phytohemagglutinin

PWM

Pokeweed mitogen

HLA

Human leukocyte antigen

PPD

Purified protein derivative

NK

Natural killer

Introduction

Mendelian susceptibility to mycobacterial disease (MSMD; MIM 209950) is a primary immunodeficiency defined by severe clinical infections with weakly virulent mycobacterial species such as nontuberculous mycobacteria [1] and bacillus Calmette–Guérin vaccines [2].

Molecular investigations to find the cause for MSMD have identified mutations in six genes involved in the interleukin (IL)-12/IL-23 dependent interferon gamma (IFN-γ) axis, highlighting the importance of this pathway in the immunity to mycobacteria. X-linked MSMD is caused by mutations in nuclear factor-kB-essential modulator (NEMO) [3] while the other five are autosomal genes. Two involved in control of the production of IFN-γ: IL12B, encoding the p40 subunit of IL12 and IL23, and IL12RB1, encoding the β1 chain of the IL12 and IL23 receptors (IL12RB1). The three other autosomal MSMD-causing genes are involved in control of the response to IFN-γ: IFNGR1 and IFNGR2, encoding the IFN-γ receptor (IFNGR) chains, and STAT1, encoding the transcription factor signal transducer and activator of transcription-1 (STAT1) [4].

The allelic heterogeneity and the varying penetrance of the inherited disorders reflect the complex interplay between host and environmental factors in the course of the disease. Complete IFNGR1 and IFNGR2 deficiencies have a poor prognosis with the development of disseminated infection in early childhood and progressively fatal disease (reviewed in [5, 6]). The other defects with mutations in IFNGR1, STAT1, IL12B, and IL12B1 are associated with residual IFN-γ-mediated immunity with varying penetrance from no mycobacterial disease, identified by screening of family members, to disseminated mycobacterial disease [5, 6]. Other IFNGR1 or IFNGR2 mutations resulting in partial receptor deficiency are associated with milder phenotype and response to IFN-γ treatment [5, 6].

The binding of IFN-γ to IFNGR1 leads to the activation of constitutively associated Janus kinases 1 and 2 (JAK1 and JAK2), which then phosphorylate tyrosine residues in the intracellular part of IFNGR1 (reviewed in [7]). Upon IFN-γ stimulation unphosphorylated STAT1 is recruited to IFNGR1, and STAT1 is phosphorylated at tyrosine 701 [7]. Phosphorylated STAT1 homodimers (gamma-activated factor (GAF)) binds gamma-activated sequences (GAS) present in the promoters of target genes involved in anti-mycobacterial immunity [8].

Following interferon alpha (IFN-α) stimulation, STAT2 is recruited to the phosphorylated IFN-α receptor (IFNAR) and STAT2 is subsequently phosphorylated by JAK1 and Tyrosine kinase 2. This leads to the STAT2 mediated recruitment of STAT1, which is phosphorylated at tyrosine 701. Active phosphorylated STAT1/STAT2 heterodimers interacts with IRF9 to form the transcription factor IFN-stimulated genes factor 3 (ISGF3) [9]. ISGF3 binds IFN-α sequence response elements (ISREs) in the promoters of target genes, thus conferring anti-viral immunity [9].

Heterozygous STAT1 mutations found to be dominant for GAF activation but recessive for ISGF3 activation result in MSMD [10, 11], but not viral disease, whereas complete impairment of STAT1-dependent responses to IFN-γ and IFN-α increases the susceptibility to both mycobacteria and viruses and is a very severe syndrome (OMIM 600555) [1214].

Here we report two new STAT1 mutations one maternally and one paternally inherited, which in combination result in MSMD in a 14-year-old patient.

Materials and Methods

Blood Samples, Cell Culture, and Flow Cytometry

EDTA and CPD-stabilized blood samples were collected from the patient and her parents. Peripheral blood mononuclear cells (PBMC) from CPD-stabilized blood samples were isolated using gradient centrifugation (Lymphoprep, Medinor A/S, Brøndby, Denmark). Cells were washed and resuspended in cell culture medium (RPMI 1640, (GIBCO, Invitrogen, Paisley, UK), with 20% of inactivated pooled human serum with l-glutamine) at 37°C, 5% CO2.

Cellular in vitro responses to mitogens: phytohemagglutinin (PHA; Invitrogen, Paisley, UK) and Pokeweed mitogen (PWM; Sigma, St. Louis, USA) and antigens: Mycobacterium tuberculosis (Statens Serum Institute, Copenhagen, Denmark), Candida albicans (Department of Clinical Microbiology, Aarhus, Denmark), Staphylococcus aureus (Department of Clinical Microbiology, Aarhus, Denmark), and tetanus toxoid (Statens Serum Institute, Copenhagen, Denmark) were tested by the lymphocyte transformation test. In brief, cells were incubated with mitogens or antigens in cell culture medium. Cultures were grown in triplicates for 96 and 144 h, and for the last 24 h, the cultures were labeled with 0.5 μCi H3-thymidin per well (Amersham, Buckinghamshire, UK). The cultures were then harvested and the incorporated radioactivity was quantified by liquid scintillation counting on a 1450 Microbeta Trilux counter (Perkin Elmer, Waltham, USA).

Test for complement activity was performed with the Complement System Screen kit according to the manufacturer’s protocol (Euro-Diagnostica, Malmö, Sweden).

The lymphocyte subpopulations were quantified by incubating whole blood with fluorochrome-conjugated antibodies directed against different phenotypical markers (CD14, CD45, CD2, CD5, CD20, T cell receptor (TCR)-αβ, TCR-γδ, HLA-DR, CD3, CD4, CD8, CD45RA, CD45ROCD21, CD19, CD16, CD56, CD18, CD29; BD Biosciences, San Jose, USA). A FACScan flow cytometer (Becton-Dickinson, San Jose, USA) was used for flow cytometrical assessment of phenotypical markers and acquiring 10,000 events. Three color fluorescence was detected using log amplification and live compensation for spectral overlap.

To analyze the cellular response to mycobacterial antigen PBMCs from the patient and a healthy control was incubated with or without inactivated mycobacterial antigen for 144 h. After incubation HLA-DR, CD54, and CD4 antibodies (BD Biosciences, San Jose, USA) were added and cells were analyzed by flow cytometry.

RNA and DNA Purification

RNA was isolated from 200-μL blood using MagNA Pure Compact RNA Isolation Kit (Roche Applied Science, Basel, Switzerland). Genomic DNA was isolated from 200-μL blood using MagNA Pure Compact Nucleic Acid Isolation Kit I (Roche Applied Science, Basel, Switzerland).

Sequencing of STAT1 mRNA and DNA

Primers are indicated in Table I (DNA technology, Aarhus, Denmark), all were used at a concentration of 0.5 μM. For sequencing of RNA reverse transcription was performed with the QuantiTect Reverse Transcription Kit according to the manufacturers protocol (Qiagen). Polymerase chain reaction (PCR) was performed using the Expand High Fidelity Polymerase (Roche Applied Science, Basel, Switzerland). PCR products were purified using the NucleoSpin Extract II kit (Macherey-Nagel, Düren, Germany) and sequenced in both directions on an ABI 3130 XL by DNA technology (DNA technology, Aarhus, Denmark). When mutations were detected, cDNA was generated and sequenced again to verify the presence of the mutation, and the mutation was verified on genomic DNA.
Table I

Primers used for sequencing of Stat1 cDNA

Name

Primer pairs (5′-3′)

Amplicon size

Stat1 cDNA F1

CTCTGTGTATATAACCTCGACA

2,437 bp

Stat1 cDNA R1

GGATGTGAAGGAACAGAGTA

Stat1 cDNA F1

CTCTGTGTATATAACCTCGACA

1,326 bp

Stat1 cDNA R2

ATTGCAGGTGCCGAAAT

Stat1 cDNA F2

GATGTGAATGAGAGAAATACAG

1,238 bp

Stat1 cDNA R1

GGATGTGAAGGAACAGAGTA

Stat1 i7

TTTGTTTCAGTGTGTCTAGTTG

628 bp

Stat1 i8

TGATCCAAGCAAGCATTG

Stat1 exon 8 R

GATCACTCTTTGCCACACCATTGG

 

Stat1 5124

GTAAATGTTAAATCTCCTAGGG

76 bp

Stat1 5199

ACAGGAGGTCATGAAAAC

High Resolution Melting Detection of STAT1 Mutation

DNA from 287 healthy voluntary blood donors and 147 patients (including the patient described in this study) referred to our department from 2008 to 2010 for analysis for suspected immunodeficiency was diluted to 10 ng/μL and analyzed on a LC480 (Roche Applied Science) with Probes Master (Roche Applied Science), LCGreeen (Idaho Technology Inc., Salt Lake City, USA) and Stat1 5124 and Stat1 5199 primers (Table I).

STAT1 Phosphorylation

Peripheral blood mononuclear cells (1 × 106 cells/mL) from the patient and a normal control were incubated for 30 min with 105 IU/mL IFN-γ (Imukin, Boehringer Ingelheim, Ingelheim, Germany), 105 IU/mL IFN-α (IntronA, Schering-Plough, Kenilworth, USA) or mock incubated. Cells were lysed in RIPA buffer (10 mM Tris–HCl (pH 8), 1% Triton X-100, 0.5% Na-deoxycholate, 150 mM NaCl, 0.2 mM phenylmethylsulfonyl fluoride, 20 μg/mL aprotinin, 10 μg/mL leupeptine, 1 mM Na-orthovanadate, 10 U/mL DNaseI, 10 μL/mL phosphatase inhibitors (Sigma, St. Louis, USA)). Cell lysates were analyzed by SDS-polyacrylamide gel electrophoresis and Western blot. STAT1 and phosphorylated STAT1 and were detected by monoclonal STAT1 and STAT1 (pY701) antibodies (BD Transduction Laboratories, San Jose, USA), respectively. Equal loading was checked by incubating the same Western blots with Lamin B (Santa Cruz, Santa Cruz, USA) Secondary ECL horseradish peroxidase-conjugated antimouse antibodies (Amersham Biosciences) were used to visualize the immunocomplexes.

Results

Case Report

A 14-year-old previously healthy girl was referred to the oncology ward at our institution with a 6 weeks history of low-grade fever, poor appetite, nausea, a weight loss of 7 kg, and bone pain. The acute-phase response was severe, with an erythrocyte sedimentation rate of 126 mm/h (<20 mm/h) and C-reactive protein 163 mg/L (<10 mg/L). She had a hemoglobin level of 5.6 g/dL and leucocytes were 18 × 109/L. Chest X-ray and high-resolution CT scan of the lungs were both normal. The PPD tuberculin test was negative, and she was HIV negative. The patient had not been exposed to BCG vaccine; she had been hospitalized at the age of three with pneumonia and was subject to tonsillectomy at the age of 11. She had serum antibodies to Epstein–Barr virus, Cytomegalovirus and Varicella zoster virus, while analysis for Herpes simplex virus antibodies was negative.

The MRI scanning showed multiple osteolytic processes in the calvarium, the right collum and corpus humeri, the right scapula, left cavitas glenoidalis, thorax skeleton, abdomen, pelvis, columna lumbosacralis, and femora. Lymphadenitis was found in collum, mediastinum, liver, retroperitoneum, and the right side of pelvis. Malignant disease was excluded.

Bronchial alveolar lavage, blood cultures, and gastric washings were negative for acid-fast organisms and no bacteria were cultured. PCR for mycobacteria was negative. Mycobacterium avium was subsequently cultured from a bone lesion. She was successfully treated with clarithromycin and rifampicin for 16 months with etambutol added for the first 2 months.

Immunological Assessment of the Patient

Before the patient was treated for the Mycobacterium avium infection, she was referred to the Department of Clinical Immunology. A lymphocyte transformation test showed a normal response after stimulation with mitogens PHA (T cell stimulation) and PWM (T cell-dependent B cell stimulation), as well as a normal response after stimulation with bacterial and fungoid antigens. The response to tetanus toxoid was negative, which may be explained by a lack of boost vaccination of the patient.

The lymphocyte subpopulations were almost normal (Table II). There was a slight lymphocytosis and eosinophilocytes were marginally increased. The absolute numbers of T cells were slightly increased, corresponding to the lymphocytosis. The CD4+ T cell count was normal, while the fraction was reduced. The absolute numbers and fraction of CD8+ T cells was increased exhibiting an overweight of CD45RO+ cells. The patient had a normal B cell count and a small increase in NK-cells. The fraction of B cells was consequently slightly decreased and the fraction of NK-cells was above normal (Table II).
Table II

Lymphocyte subpopulations

 

Cell count (109/L)

Normal rangeb

Relative distribution (%)

Normal rangeb

Leukocytes

9.5

5.1–7.3

  

Lymphocytes

4.0

1.9–2.9

42

35–45

T lymphocytes (totally)

2.7

1.4–2.2

66a

 

T cell receptor αβ

  

62a

35–77

T cell receptor γδ

  

4.9a

2–17

Activated T lymphocytes

0.05

0.07–0.16

1.3

5–9.5

T helper cells

1.0

0.6–1.2

25.5a

33–45

 Naïve T helper cells

  

31

44–73

 Memory T helper cells

  

64

12–42

Cytotoxic T cells

1.7

0.6–0.9

42a

30–36

 Naïve cytotoxic T cells

  

40

47–80

 Memory cytotoxic T cells

  

54

4–20

B lymphocytes

0.4

0.26–0.51

10.6a

13–19

NK-cells

0.7

0.18–0.34

17a

7–15

aPercent of lymphocyte gate

bNormal range according to [18]

The analysis for complement activity showed normal activity of the classical, the alternative, and the lectin activation pathways, as well as normal MBL activity and concentration.

Immunoreactivity to Mycobacterial Antigen

The expression of HLA-DR and CD54 in monocytes is known to be up regulated after cellular exposure to mycobacterial antigen. Cells were stimulated with inactivated mycobacterial antigen for 144 h, and cells were subsequently labeled with fluorochrome-conjugated antibodies against HLA-DR, CD54, and CD4. An antigen dosage-dependent up-regulation of the surface expression of both HLA-DR and CD54 in monocytes of 72% and 34%, respectively, was detected by flow cytometry. This corresponds to the up-regulation seen in control cells. In CD4+ T cells, an adequate and antigen dosage-dependent response to mycobacterial antigen was detected as well. The fraction of HLA-DR/CD54 double-positive cells increased from 3.3% to 10.1%.

Genetic Analysis of Genes Involved in MSMD

Initially, the exons of IFNGR1 and IFNGR2 as well as the coding area of IL12RB1 mRNA and IL12B mRNA were sequenced without finding any mutations. In contrast mutations were found in STAT1 mRNA.

One mutation was detected in exon 4 of both the patient and her father (Fig. 1a). The mutation is a transition from G to A at position 5151 (G5151A), resulting in an amino acid change from alanine to threonine at position 46 (Ala46Thr) (Fig. 1b, genbank accession number: GU211347). We tested whether this mutation was a single nucleotide polymorphism present in the Danish population. The presence of a G to A transition will lower the melting properties of DNA, and this was exploited in a real time PCR analysis using high-resolution melting (Fig. 2). We thus analyzed 287 healthy blood donors and none of these healthy controls harbored the mutation. We furthermore tested whether this mutation could be found in a group of patients referred to our department on clinical suspicion of immunodeficiency. None of these patients, except the patient described in this study had this mutation. Thus the mutation is rare, the 95% confidence interval for allele frequency is 0–0.007 (Bayesian calculation).
https://static-content.springer.com/image/art%3A10.1007%2Fs10875-010-9480-8/MediaObjects/10875_2010_9480_Fig1_HTML.gif
Fig. 1

a STAT1 genotype and phenotype of the analyzed family. The mother and father both carry a STAT1 mutation while being healthy and the patient inherited a mutated allele from both parents. b cDNA sequences in the sense orientation of part of exon 4 from the patient and her parents. The mutation is present on the allele with exon 8, as shown by the sequence generated with a reverse primer situated in exon 8. c The human STAT1 coding region with its known pathogenic mutations. The coiled-coil domain (CC), DNA binding domain (DNA-B), linker domain (L), SH2 domain (SH2), tail segment domain (TS), and trans-activator domain (TA). Mutations shown in red are recessive mutations associated with complete STAT1 deficiency and subsequent impaired GAF and ISGF3 activation. Mutations shown in blue result in partial STAT1 deficiency in heterozygous individuals, impaired IFN-γ-induced GAS-binding activity, and normal IFN-α-induced ISRE-binding activity. Mutations found in this study are indicated in green

https://static-content.springer.com/image/art%3A10.1007%2Fs10875-010-9480-8/MediaObjects/10875_2010_9480_Fig2_HTML.gif
Fig. 2

Melting peak analysis of a 75 bp PCR fragment (STAT1 5124-5199) encompassing the G5151A mutation. Presence of the mutation alters the melting properties of the PCR fragment by lowering melting temperature. The arrow at A points to the curves for the patient and her mother, while arrow at B shows the curve for the father and 20 healthy donors

The other expressed STAT1 allele in the patient is a splice variant which lacks exon 8 (genbank accession number: GU211348). We identified the allele by sequencing cDNA, and initially found that the sequence was scrambled at the transition to exon 8 both in the forward and reverse sequencing direction in cDNA from both the patient and her mother. When employing a reverse sequencing primer situated in exon 8 (Stat1 exon 8 R, Table II), only the G5151A allele was seen in the patient (Fig. 1b), and only the wt allele was seen in the mother as expected. The splice variant is the result of a splice site mutation in exon 8, at position 16034 (A16034G, genbank accession number: HQ284032). This allele, which is inherited from the mother, may produce a truncated protein of 194 amino acids compared with full-length STAT1, which is 750 amino acids (Fig. 1c).

STAT1 Phosphorylation

To test whether the observed mutations influence the STAT1 protein, the phosphorylation response to IFN-α and IFN-γ was analyzed. STAT1 expression was clearly lower in PBMC from the patient than in PBMC from a healthy control as shown by Western blotting (Fig. 3). Phosphorylation at tyrosine 701 was reduced after IFN-γ stimulation in the patient. Phosphorylation after IFN-α treatment in the patient was lower than in the healthy control, but this may be due to the reduced STAT1 expression (Fig. 3). Equal loading was checked by reprobing the blots with lamin B.
https://static-content.springer.com/image/art%3A10.1007%2Fs10875-010-9480-8/MediaObjects/10875_2010_9480_Fig3_HTML.gif
Fig. 3

Reduced STAT1 expression and reduced IFN-γ induced phosphorylation of STAT1. Western blot of total protein extract (10 μg) from PBMC from the patient and a healthy control probed with specific antibodies against STAT1, phosphorylated STAT1 (pY701) and lamin B. PBMCs were not stimulated (–) or stimulated for 30 min with 105 IU/mL IFN-γ (γ) or 105 IU/mL IFN-α (α)

Discussion

The patient described in this study initially did not display immunological defects that would explain her serious condition. The discovery of the Mycobacterium avium infection led to a search for a defect in the genes known to be associated with MSMD.

We thus describe two novel mutations in STAT1, which in combination result in partial STAT1 deficiency and MSMD. One expressed allele of the patient lacked exon 8, while the other allele had an amino acid substitution A46T. The previously described STAT1 deficiencies associated with this syndrome are situated in completely different regions of STAT1 (Fig. 1c). The L706S STAT1 mutation is located in the tail segment domain in close vicinity to tyrosine 701, and results in defects in phosphorylation at Y701, while expression of STAT1 is normal [10]. The E320Q and Q463H mutations are situated in the DNA binding domain, and consequently result in impaired GAS binding but normal Y701 phosphorylation and normal STAT1 expression [11]. In this study we describe two mutations. The A46T mutation found in the patient and her father is situated in the amino (N)-terminal region. This region plays a role in stabilizing STAT dimer–dimer interactions [15] and is required for interaction between STAT1 and the transcriptional coactivator protein CBP [16]. The N-terminal domain has a well-defined hydrophobic core that is highly conserved across the STATs, consistent with a stable and well defined fold [17].

The alanine 46 residue is situated in α-helix 5 and is one of the conserved residues in the hydrophobic core, and since threonine is hydrophilic this may disrupt the folding of the N-terminal region. The other allele is an alternatively spliced form of STAT1 without exon 8. This allele may only give rise to a much shortened protein which only retains an intact N-terminal region (Fig. 2). If expressed, this protein is nonfunctional at best and in a homozygous form it would most certainly result in lethal viral disease.

The mutations seem to have an impact on several functions of the protein. The two alleles may be primarily deleterious by impairing expression of STAT1, however additional impacts may occur. The two mutations collectively may impair N-terminal STAT1-dependent phenomenons, such as homo-dimerization (STAT1/STAT1), heterodimerization (STAT1/CBP, STAT1/P48) and phosphorylation. Subsequently, STAT1 signaling may be influenced, and we observed a reduced expression of STAT1 as well as decreased Tyrosine 701 phosphorylation of STAT1. As a consequence GAF and ISGF3 activation after IFN-γ and IFN-α stimulation, and downstream ISGs induction may be impaired, and the patient may be at risk for viral diseases as well, although she did not display increased susceptibility to viral disease so far.

We found that cells from the patient did up regulate CD54 and HLA-DR in response to mycobacterial antigen. This shows, as expected, that the cellular response to mycobacterial antigen is not abrupted by the mutations found in STAT1. This correlates with the fact that the patient had no earlier episodes of mycobacterial disease, and had lived 14 years before onset.

The patient remains susceptible to mycobacterial infections. Importantly, as we find that IFN-γ does to some extent result in phosphorylation of STAT1, IFN-γ treatment would potentially reduce the severity of a new episode of mycobacterial infection, however she is currently healthy and without any prophylactic treatment.

Copyright information

© Springer Science+Business Media, LLC 2010