Why we can’t use LTTs, yet

A line of T cells (called Ob.2F3) expressing the same T cell receptor (TCR) from an MS patient was studied in 2014 and it was found to proliferate when incubated with 4824 different peptides. Thirty-three of them were further studied (see figure) and found to belong to both Homo sapiens and several different, unrelated microbes (Birnbaum ME et al. 2014). The taking home message here is that T cells are not specific to a single pathogen, they are highly cross-reactive, as it was already pointed out in this pivotal study: (Mason DA 1998). And this means that we can’t use lymphocyte transformation tests (LTTs) the way we do now. 

I feel really frustrated when patients send me their LTTs and ask me to comment the results. I have to say that they have vasted their money and that these results are useless. I do hope that my blog can make a difference and stop this unfair commerce at the expenses of desperate folks.

Crossreactive epitopes
Figure. A set of 33 peptides (both human and environmental) predicted to be specific epitopes for both Ob.1A12 and Ob.2F3. From (Birnbaum ME et al. 2014).






Test a due livelli per la malattia di Lyme

Per ottimizzare i test sierologici in termini di sensibilità e specificità, si adotta in Europa, quanto negli Stati Uniti, un algoritmo a due livelli (vedi Figura 1). Il primo livello è costituito da un test ad alta sensibilità (il test ELISA); coloro che sono negativi a questo test vengono considerati sieronegativi, coloro che sono positivi accedono a un test di secondo livello altamente specifico (il western blot), il cui scopo è quello di minimizzare il numero di falsi positivi (Wilske B. et al. 2007).

two tiers.png
Figura 1.  Algoritmo decisionale per il test a due livelli.

In Tabella 1 sono riportati i dati statistici relativi a un test a due livelli in cui il primo livello è rappresentato da un test ELISA con antigeni ottenuti da sonicato da cellula intera di Borrelia afzelii per le IgM, con aggiunta di VlsE ricombinate da B. afzelii, B. garinii e B. burgdorferi ss per le IgG; il secondo livello è invece costituito da un western blot i cui antigeni sono proteine estratte da B. burgdorferi ss (ceppo B31) e da Borrelia afzelii (ceppo PKo) con aggiunta di VlsE ricombinate da B burgdorferi ss (ceppo B31) (Branda JA. et al. 2013). Si può notare che con questa procedura si ha un 12% di falsi negativi nel gruppo neuroborreliosi e una percentuale di falsi positivi sempre minore del 2%.

doppio livello.png
Tabella 1. Sono riportati sensibilità, specificità, valore predittivo positivo (VPP) e valore predittivo negativo (VPN) per un western blot europeo. EM sta per eritema migrante, NB per neuroborreliosi, ACA per acrodermatite cronica atrofica. IgG e IgM sono considerate insieme. Da (Branda JA. et al. 2013) con modifiche.

A conclusione di questo paragrafo, si segnala che nel 2011 i CDC hanno modificato le loro raccomandazioni, includendo la possibilità dell’uso diretto di un immunoblot per le IgG, senza passare per il test di primo livello (R).


Considera una donazione per sostenere questo blog.



False positive IgM tests in infectious and autoimmune diseases

False positive IgM tests in infectious and autoimmune diseases


IgM tests present a high rate of false positive results. This can lead to misdiagnosis, inappropriate treatments, and lack of treatment for the true aetiology (Landry ML. 2016). In the following table, I have collected several well-documented cases of IgM tests falsely positive for an infectious disease. In most of these studies, the cause of the false positive result was found to be either another acute infection or autoantibodies, including rheumatoid factor (RF), an IgM that binds to the Fc region of IgG. So it might be important to determine the exact origin of a false positive IgM test, in cases where a diagnosis is hard to find: it could be the clue that ultimately leads to the true aetiology.

IgM falsely positive for: True aetiology N. of cases Reference
Hantavirus IgM

Nombre virus IgM

Adenovirus 1 Landry ML. 2016
Measles virus IgM Sulfa drug allergy 1 Landry ML. 2016
HAV IgM CHF 1 Landry ML. 2016
HEV IgM HAV (IgM+) 1 Landry ML. 2016
HSV IgM VZV (IgM+) 1 Kinno R. et al. 2015
VZV (IgM+) 11/50 Ziegler T. et al. 1989
Parvovirus B19 (acute) 5/65 Costa E. et al. 2009
RF 9/50 Ziegler T. et al. 1989
RF 1 Pan J. et al. 2018
Anti-HDF 2/50 Ziegler T. et al. 1989
HSV-2 IgM HSV-1 1 Landry ML. 2016
VZV Anti-HDF 5/74 Ziegler T. et al. 1989
HSV (IgM+) 8/54 Ziegler T. et al. 1989
RF 8/54 Ziegler T. et al. 1989
EBV VCA IgM CMV (IgM+) 1 Landry ML. 2016
CMV (IgM+) 7/50 Aalto MS et al. 1998
B. burgdorferi (IgM+) 2 Pavletic A. Marques AR. 2017
HEV (IgM+) 33,3% Hyams C et al. 2012
CMV IgM HEV (IgM+) 24,2% Hyams C et al. 2012
Anti-HDF 10/75 Ziegler T. et al. 1989
RF 3/75 Ziegler T. et al. 1989
WNV IgM HSV-2 1 Landry ML. 2016
B. burgdorferi IgM

(OspC and/or BmpA)

HSV 2 (IgM+) 1 Strasfeld L. et al. 2005
VZV (acute) 5/12 Feder HM. et al. 1991
EBV (acute) 14/58 Goossens HA. et al. 1998
CMV (acute) 13/58 Goossens HA. et al. 1998
Mycoplasma IgM WNV (IgM+) 1 Landry ML. 2016

List of abbreviations. CHF, congestive heart failure; CHIK virus, Chikungunya virus; CMV, cytomegalovirus; EBV, Epstein-Barr virus; HAV, hepatitis A virus; HDF, human diploid fibroblast cells; HEV, hepatitis E virus; HIV, human immunodeficiency virus; HHV-6, human herpesvirus type 6; HSV, herpes simplex virus; RF, rheumatoid factor; VZV, varicella-zoster virus; WNV, West Nile virus.

Cross-reactivity with other pathogens

One possible cause for false positive results is cross-reactivity between antigens that belong to different pathogens. A little-known example of this phenomenon comes from the research on ME/CFS: in the study that ultimately ruled out the involvement of XMR virus in the pathogenesis of ME/CFS, antibodies to that pathogen were found in about 6% of both cases and healthy controls, whereas the molecular testing turned out to be negative in all participants (Alter HJ. et al. 2012). So, sera reactivity to XMRV is likely due to a relatively common pathogen that has an antigen similar to another one belonging to XMRV.

In one case of false positive HSV IgM due to VZV infection, the serum/CSF IgM ratio as a function of time had the same profile for both the viruses, suggesting cross-reactivity (Kinno R. et al. 2015). Cross-reactivity between HSV IgM and VZV IgM seems quite common, with both false positive HSV samples due to reactivity to VZV and false positive VZV IgMs due to IgM against HSV (Ziegler T. et al. 1989).

Interestingly enough, although OspC is considered to be a highly specific antigen of B. burgdorferi, OspC IgM is often positive in patients with active EBV or CMV infections (Goossens HA. et al. 1998). If cross-reactivity was responsible for false positive OspC IgM in infectious mononucleosis, we would expect false positive IgM for EBV and CMV in early Lyme disease. And this is is exactly what has been found in two cases of acute Lyme disease, where falsely positive IgM to VCA has been documented (Pavletic A. Marques AR. 2017).

Latent infections reactivation

It has been described a rise of EBV VCA IgM titers in CMV primary infections. This is likely due to EBV reactivation in many cases. This can lead to a misdiagnosis of a primary EBV infection, instead of a primary CMV infection. This error can have serious consequences during immune suppression or pregnancy, when CMV infections are health-threatening (Aalto MS et al. 1998).

Rheumatoid factor interference

As mentioned in the introduction, rheumatoid factor (RF) is an autoantibody – mainly of the IgM subclass – that is found in most of the patients with rheumatoid arthritis (Hermann E. et al. 1986). It binds the constant region (Fc region) of human IgGs and thus can bind the enzyme-linked immunoglobulins often used in serologic assays, leading to falsely positive results (Pan J. et al. 2018).

Cross-reactivity with autoantigens

Another possible cause of falsely positive IgM tests to a pathogen is the presence of autoantibodies other than RF. Autoantibodies to fibroblast cells have been found to be the cause of a false positive IgM test for HSV and CMV (Ziegler T. et al. 1989). This kind of reactivity to self-antigens is probably non-specific of a particular autoimmune disease, and it has been found for instance in pretibial myxedema, Graves’ disease, and Hashimoto’s thyroiditis (Arnold K. et al. 1995).

The effect of prevalence on the rate of false positive results

The likelihood of a false positive result is inversely correlated with the prevalence of the pathogen in the specific population considered. In other words, the rarer the disease, the more likely a positive test for that disease is a false positive. This can be easily seen introducing the predictive positive value (PPV), which is the probability that a positive test is really a true positive (Lalkhem AG. et McCluskey A. 2008). PPV is given by


In the following figure, you can see how PPV increases as the prevalence increases. This diagram has been plotted considering a sensitivity of 67% and a specificity of 53%.

PPV 2.png



Epitope mapping in Lyme disease

Epitope mapping in Lyme disease


A new study from Columbia University, Stony Brook University and the CDC – that has seen the collaboration of Brian Fallon and W. Ian Lipkin, among others – has proposed a new serological test for Lyme disease and coinfections (Babesia microti, Anaplasma phagocytophilum, Ehrlichia chaffeensis, Rickettsia ricketsii and some viruses that are not present in Europe) (Tokarz R et al. 2018). Current serologic analyses consist of a two-tiered algorithm:  a measure of serum activity against a main immunogenic protein of B. burgdorferi (whether it is VlsE or its peptide C6) is followed by a western blot, where the immune response to a set of several full-length proteins is performed. This method lacks sensitivity for early Lyme disease (Aguero-Rosenfeld ME et al 2005) and – according to studies on the animal model of Lyme disease – it might miss some cases of disseminated infection too (Embers ME, 2012), (Nicholas A, 2017), (Embers ME 2017). So, there is common agreement that a better test is urgently needed.

The quest for immunogenic peptides in Lyme patients: the set of proteins

Most B cells epitopes on non-denaturated proteins (i.e. proteins that conserve their tertiary structure) are believed to be conformational (Morris, 2007) but it is also true that in the average B cell epitope, a linear stretch of 5 amino acids is reported (Kringelum, et al., 2013). This means that it is conceivable to search for new immunogenic peptides in Lyme disease with the following method: each protein from Borrelia, known to be immunogenic in humans, is divided in peptides with a fixed length, then each of these peptides is exposed to sera from patients. Those peptides that strongly bind sera from patients are eligible as immunogenic peptides useful for diagnostic purposes. This is exactly what has been done in this study, and this elaborate analysis has been performed not only for Borrelia burgdorferi but also for the other tick-borne pathogens already mentioned. In particular, the length of the peptides has been fixed to be 12, and contiguous peptides have an overlapping of 11 amino acids. The set of immunogenic proteins chosen for each pathogen are reported in table 1.

table 1
Table 1. Set of proteins from various pathogen chosen by the Authors.

A new array of diagnostic peptides in Lyme disease

The analysis was conducted using sera from 66 Lyme patients (27 early Lyme, 19 with positive IgG Western blot, 10 with acute neuroborreliosis). The end result is the set of peptides reported in table 2. There is also a set of peptides specific for neuroborreliosis (table 3).

table 2.png
Table 2. Peptides on the second column of this table are specific immunogenic proteins in Lyme disease patients. In the first column proteins they belong to are collected.
table 3.png
Table 3. Imunogenic peptides in neuroborreliosis.

Flagellin B: a new specific peptide for diagnostic purposes is born

It is widely recognized that reactivity of sera to Flagellin B (FlaB, p41) is common in healthy persons too (Chandra A et al. 2011). This is because this protein is highly conserved among bacteria, so there is cross-reactivity and FlaB is almost without utility for diagnostic purposes. That said, one of the results of this study is that peptide 211-223 fro FlaB is highly immunogenic in patients and is also not-cross-reactive with flagellar proteins from other bacteria (I have checked using this peptide as query sequence in a BLAST search among bacteria, with a word of 3 and with standard settings: no match has been found). So we have a new specific peptide for diagnostic purposes. In figure 1 this epitope is reported in yellow, on the 3D structure of Flagellin B.

Figure 1. Peptide 211-223 of flagellin B, in yellow. This structure is built by Modbase using the sequence P11089 and using flagellin of E. coli as a template.

The strange case of VlsE and C6

Vmp-like sequence expressed (VlsE) is a major immunogenic protein of Borrelia burgodorferi, widely used for the first-tier test of Lyme disease (Aguero-Rosenfeld ME et al 2005). This protein goes through a process of variation during mammalian infection that involves six variable regions (VR1-VR6) and it has been postulated that this gene recombination of the locus that codes for VlsE is responsible – at least in part – for the ability of B. burgodorferi of evading the immune system (Zhang JR et al. 1997). Nevertheless, six regions of this protein are invariable (IR1-6) and these ones have been studied as possible peptides to use for diagnosis. One of them, peptide IR6 (also called C6), has been used as a diagnostic tool (Liang FT et al. 1999). In the present study, the peptide that pops up from the analysis is the same C6 (peptide 274-290 in table 1), while in neuroborreliosis there are two more peptides on VlsE that appear to be useful markers. I have reported these three peptides on the 3D structure of VlsE, the one experimentally determined in (Eicken C et al. 2002), see figure 2. It is worth noting now that peptide 274-290 (C6) is buried inside the protein, so it is difficult to understand how it could be a B cell epitope since B cell epitope protrudes from the surface (Thornton, et al., 1986) and this is probably because B cell receptors need to bind their specific antigen, for B cell to be activated. This is the very first time that I realize that C6 is not surface exposed. So, how could the immune system generate an antibody to a hidden peptide? I have to think about that.

Figure 2. Immunogenic peptides on VlsE (left) and on OspC (right).

OspC and the buried epitope

Another weird case is that of OspC, where one of the two main epitopes found in patients with neuroborreliosis is buried in the cytoplasmic membrane of Borrelia burgdorferi (figure 2, right). This is an uncommon case, I guess, for the same reason mentioned above: BCRs need surface exposure to bind their specific epitope. The other peptide (132-144) appears to be a classic surface-exposed B cell epitope.





Mark Davis and the search for the universal immune test

Mark Davis and the search for the universal immune test

A traslation of this blog post to Spanish can be downloaded here. I would like to thank Humbert.Cat for the translation.

1. Introduction

These are some notes about the talk that Mark Davis gave during the Community Symposium held in August at Stanford (video). I will introduce some basic notions about T cell receptors (TCR) in paragraphs 2, 3, 4, and 5. Paragraphs 6 is a description of an innovative technology developed by Mark Davis and his colleagues, based on information gathered from the video itself and three research papers published by Davis and others in the last 4 years. This background should be hopefully enough to allow a good understanding of the exciting pilot data presented by Mark Davis on T cell activity in ME/CFS (paragraph 7), and in chronic Lyme (paragraph 8), and to realize why this technology promises to be some sort of universal test for any kind of infectious and autoimmune diseases, known or unknown.

2. T cells

T cells are a type of leukocytes (also known as white blood cells), the cellular component of our immune system. Most of our circulating T cells are represented by T helper cells (Th cells) and cytotoxic T lymphocytes (CTL). While the function of Th cells is to regulate the activity of other leukocytes through the production of a wide range of chemicals (cytokines), CTLs are directly involved in the killing of host cells infected by pathogens. T cells belong to the adaptive arm of the immune system, along with B cells (the factories of antibodies), and as such, they are meant to provide a defence tailored to specific pathogens: our immune system can provide not only antibodies specific for a given pathogen but also specific T cells (both Th cells and CTLs). The innate arm of the immune system (which includes natural killer cells, macrophages, dendritic cells, mast cells…) on the other hand can provide only a one-fits-all type of defense, which represents the first line of immune response, during an infection.

3. T cell receptor

T cells search for their specific pathogens thanks to a molecule expressed on their surface, called T cell receptor (TCR). In figure 1 you can see a schematic representation of the TCR and of the mechanism by which T cells recognize their targets. Antigens (proteins) from pathogens are presented to T cells by other cells of our body: they are displayed on molecules called major histocompatibility complex (MHC), expressed on the outer membrane; if the antigen fits the TCR of a specific T cell, then this T cell is activated and proliferates (clonal expansion). The two chains (α and β) are assembled using the transcription of gene segments with several copies each: in other words, TCRs are assembled with peptides chosen randomly within a set of several possible choices. This leads to a repertoire of 10^15 possible different TCRs (Mason DA 1998). Each T cell displays only one type of TCR.

Figure 1. Upper half. Th cells and CTLs share the same TCR: in both cases this molecule is the assembly of two peptides (chain α and chain β), but while the TCR of Th cells (on the right) is expressed next to the molecule CD4 (which binds to class II MHC), the TCR of CTL is associated with the molecule CD8 (on the left), which is specific for MHC I. Black bars represent four chains (a complex called CD3) that are involved in the signaling of the TCR with the nucleus of the cell (by Paolo Maccallini). Lower half. A beautiful structural representation of the TCR, bound to the peptide-MHC complex (pMHC), from (Gonzàlez PA et al. 2013). In green the peptide, in blue the β chain, in dark green the α chain. CDRs (complementarity determining regions, orange) are composed of those residues of the α and β chains that directly bind the pMHC.

4. T helper cells

Th cells can recognize only antigens presented by class II MHC: this class of MHC is expressed on the outer membrane of some leukocytes, mainly dendritic cells, B cells, and macrophages (referred to as antigen presenting cells, APCs). MHC II engages the TCR of Th cells thanks to peptide CD4 (expressed exclusively by Th cells). The antigen presented by MHC II is a peptide with a length of 13-17 amino acids (Rudensky, et al., 1991) (figure 2).

Figure 2. The TCR expressed by a Th cell binds an epitope presented by a class II MHC expressed on the plasma membrane of an APC. Chains α and β of MHC II are also represented (by Paolo Maccallini).

5. Cytotoxic T lymphocytes

TCRs expressed by CTLs can bind only antigens displayed by class I MHC, which can be found on the outer membrane of any cell of our body. CD8 is the molecule that makes the TCR expressed by CTLs specific for MHC I. While antigens presented by APCs belongs to pathogens that have been collected on the battlefield of the infection, peptides displayed by class I MHC of a specific cell belong to pathogens that have entered the cell itself, therefore they are the proof of an ongoing intracellular infection (figure 3). When a CTL recognizes an antigen that fits its TCR, then the CTL induces apoptosis (programmed death) of the cell that displays it. Antigens presented by MHC I are peptides in the range of 8 to 10 amino acids (Stern, et al., 1994).

Figure 3. An infected cell displays a viral antigen on its MHC I. The TCR of a CTL binds this peptide and send a signal to the nucleus of the CTL itself, that responds with the induction of apoptosis (releasing granzymes, for instance) of the infected cell (by Paolo Maccallini).  

6. The universal immune testing

In his talk, Mark Davis presents an overview of some basic concepts about the immune system, before introducing his exciting new data about ME/CFS and post-treatment Lyme disease syndrome (PTLDS, also known as chronic Lyme). But he also describes a few details of a complex new assay that is theoretically able to read all the information packed in the repertoire of TCRs present – in a given moment – in the blood of a human being. As such, this test – that I have named the universal immune testing – seems to have the potential to determine if a given patient has an ongoing infection (and the exact pathogen) or an autoimmune disease (and the exact autoantigen, i.e. the tissue attached by the immune system). To my understanding, this assay requires three steps, described in the following sections.

6.1. First step: TCR sequencing

As said in paragraph 3, when T cells encounter their specific peptide presented by MHC, they proliferate so that in blood of patients with an ongoing infection (or with a reaction against self, i.e. autoimmunity) we can find several copies of T cells expressing the same TCR: while in healthy controls about 10% of total CD8 T cells is represented by clones of a few different T cells (figure 4, first line), in early Lyme disease – an example of active infection – and in multiple sclerosis (MS) – an example of autoimmune disease – we have a massive clonation of a few lines of CTLs (figure 5, second and third line, respectively). The first step of the universal immune testing is represented by the identification of the exact sequence of TCRs expressed by T cells in blood, as reported in (Han A et al. 2014) where it is described how to sequence genes for the α and the β chain of a given T cell. This approach allows to build graphs of the kind in figure 4, and therefore to determine whether the patient has an abnormal ongoing T cell activity or not. If a clonal expansion is found, then we can speculate that either an active infection is present or some sort of autoimmune condition.

Clonal expansion CD8.png
Figure 4. Each circle represents a patient. In the first line, we have four healthy controls, with no clonal expansion of CD8 T cells (the first one, left) or with only a low-level clonal expansion (slices in blue, white, and grey). In the second line, we have four patients with active Lyme disease (early Lyme) and all of them present a massive expansion of only three different T cells (slices in red, blue and orange). In the third line, we have four MS patient with most of their CD8 T cells represented by clones of a bunch of T cells. From the talk by Mark Davis.

6.2. Second step: TCR clustering

Mark Davis and his group have been able to code a software that allows to identify TCRs that share the same antigen, either within an individual or across different donors. This algorithm has been termed GLIPH (grouping of lymphocyte interaction by paratope hotspots) and has been found capable – for instance – to cluster T CD4 cell receptors from 22 subjects with latent M. tuberculosis infection into 16 distinct groups, each of which comprises TCRs from at least 3 different donors (Glanville J et al. 2017). Five of these groups are reported in figure 5. The idea here is that TCRs that belong to the same cluster, react to the same peptide-MHC complex (pMHC).

Figure 5. Five group of TCRs from 22 different donors with latent tuberculosis, clustered by GLIPH. The first column on the left has TCRs IDs, the second one reports donors IDs. Complementarity determining regions (CDR) for the β and the α chains are reported in the third and fifth column, respectively. From (Glanville J et al. 2017).

6.3. Third step: quest for the epitope(s)

As we have seen, this new technology allows to recognize if T cell clonal expansion is an issue in a given patient, by sequencing TCRs from his peripheral blood. It also allows to cluster TCRs either within an individual or across different patients. The next step is to identify what kind of antigen(s) each cluster of TCRs reacts to. In fact, if we could recognize these antigens in a group of patients with similar symptoms, with T cell clonal expansion and similar TCRs, we would be able to understand whether their immune system is fighting a pathogen (and to identify the pathogen) or if it is attacking host tissues and, if that was the case, to identify what tissue. As mentioned, the number of possible TCR gene rearrangement is supposed to be about 10^15, but the number of possible Th cell epitope is about 20^15 which is more than 10^19. This implies that TCRs have to be cross-reactive to some extent, in order to recognize all possible peptides presented by MHCs (Mason DA 1998). The exact extent of this cross-reactivity and the mechanism by which it is obtained has been elucidated by Mark Davis and his colleagues in a recent paper (Birnbaum ME et al. 2014) that gives us the third step of the universal immune testing. The aim of this phase is to take a given TCR and to find the repertoire of his specific antigens (as said, one TCR reacts to several antigens). In order to understand how this is possible let’s consider one of the experiments described in the paper mentioned above. The researchers considered two well-defined TCRs (named Ob.1A12 and Ob.2F3), cloned from a patient with MS and known to recognize peptide 85-99 (figure 6) of myelin basic protein (MBP) presented by HLA-DR15. They then prepared a set of yeast cells expressing HLA-DR15 molecules, each presenting a different peptide of 14 amino acids, with fixed residues only at position 1 and 4, where the peptide is anchored to MHC (figure 6, left). When copies of Ob.1A12 are added to this culture of yeast cells expressing pMHC complexes, they bind only some of them, and as you can see in the right half of figure 6, for each position of the epitopes bound by Ob.1A12, there is an amino acid that is more likely: for instance, the typical epitope of Ob.1A12 preferentially has alanine (A) at position -4, histidine (H) at position -3, arginine (R) at position -2, and so forth. As you can see, histidine (H) at position 2 and phenylalanine (F) at position 3 are obligate amino acids for a Ob.1A12 epitope.

Figure 6. On the left: peptide 85-99 of myelin basic protein (first row) is known to be an epitope for Ob.1A12. At position 1 and 4 it has two residues that allow its binding to the MHC molecule. At position -2, -1, 2, 3, and 5 we find those residues that bind the TCR. The second row represents the generic epitope of the peptide library used to identify the degree of crossreactivity between all the possible Ob.1A12 specific epitopes. On the right: the likelihood of amino acids for each position of Ob.1A12 specific epitopes is represented by shades of violet. As you can see, histidine (H) at position 2 and phenylalanine (F) at position 3 are obligate amino acids for a Ob.1A12 epitope. From (Birnbaum ME et al. 2014).

The table on the right side of figure 6 is, in fact, a substitution matrix with dimension 14×20, a tool that can be used to scan the peptide database in order to find, among all the known peptides expressed by living creatures, all the possible Ob.1A12 specific epitopes. Substitution matrices are commonly used for the so-called peptide alignment, a technique that aims at the identification of similarities between peptides. These matrices are based on evolutionary considerations (Dayhoff, et al., 1978) or on the study of conserved regions in proteins (Henikoff, et al., 1992). But the search for specific epitopes of a given TCR requires (as we have seen here for Ob.1A12) a substitution matrix built ad hoc for that TCR: each TCR requires its own substitution matrix that is obtained adding clones of that TCR on a culture of yeast cells presenting a huge peptide library on their MHCs, and analyzing data from this experiment. So, quite a complex process! In the case of Ob.1A12, this process led to 2330 peptides (including MBP), while the Ob.2F3 specific substitution matrix found 4824 epitopes within the whole peptide database. These peptides included both non-human proteins (bacterial, viral…) and human peptides. For 33 of them (26 non human and 7 human proteins), this group of researchers performed a test in order to directly verify the prediction: 25/26 of environmental peptides and 6/7 of the human peptides induced proliferation of T cells expressing Ob.1A12 and/or Ob.2F3, and this is a huge proof of the validity of this analysis! These 33 peptides are reported in figure 7. This is the last step of the universal immune testing, the one that from the TCR leads to the epitopes. As you have seen, a huge set of different peptides from different sources is linked to each single TCR, in other words, crossreactivity is an intrinsic property of TCR. This also means that lymphocyte transformation tests (LTTs), widely used in Europe for the detection of infections like Borrelia burgdorferi and others, bear a high risk of false-positive results and require a process of experimental and theoretical validation, that is currently lacking (see also this post on this issue).

Crossreactive epitopes.JPG
Figure 7. A set of 33 peptides (both human and environmental) predicted to be specific epitopes for both Ob.1A12 and Ob.2F3. From (Birnbaum ME et al. 2014).

We are now ready to fully appreciate the pilot data that Mark Davis presented at the Symposium on CD8 T cell clonal expansion in ME/CFS and in chronic Lyme.

7. We have a hit!

Mark Davis, along with Jacob Glanville and José Montoya, has sequenced TCRs from the peripheral blood of 50 ME/CFS patients and 49 controls (first step of the universal immune testing, remember?), then they have clustered them using the GLIPH algorithm (second step). They have found 28 clusters with more than 2500 similar sequences each, where each cluster collects multiple sequences from the same individual as well as (which is perhaps more important) sequences from different patients (figure 8). The cluster that I have circled in red, for instance, is a collection of more than 3000 similar TCRs. The presence of this wide clusters in ME/CFS patients, compared to healthy controls, represents an indirect proof of a specific T cell response to some common trigger in this group of patients, which might be a pathogen or a tissue of the body (or both).

Clustered TCR
Figure 8. In ME/CFS, TCRs sequences from 50 patients form 28 clusters with more than 2500 similar sequences, and this is clearly not the case in healthy controls. This point to some specific immune response to a pathogen or to a human tissue (or both). This slide is from the talk by Mark Davis.

Among these 50 ME/CFS patients, Davis and colleagues selected 6 patients with similar HLA genes (figure 9, left), 5 females among them, and studied their TCRs deeper. In the right half of figure 9, you can see for each patient the degree of CTL clonal expansion. Remember that in healthy controls only about 10% of CTLs is composed by clones of a few cells (figure 4, first raw), while here we see that about 50% of all CTLs is composed by clones. So, a “marked clonal expansion” of CD8 T cells, as Davis said.

ME subjects CD8
Figure 9. On the left: 6 MECFS patients with similar HLA genes have been selected. Patient ID is reported in the first column on the left, the second column indicates the age of each patient, the third indicates the gender, the fourth column is about exposure to cytomegalovirus, the third one is on MHC I genes. On the right: analysis of clonal expansion of CD8 T cells for each of the six patients. There is a marked clonal expansion (about 50%) compared to healthy controls (about 10%).

Sequences of α and β chains of TCRs from three of the six patients (patients L4-02, L4-10, and L3-20) are reported in figure 10 (I have verified that in fact these are sequences of α and β chains of human TCRs using them as query sequences in standard protein BLAST).

TCR epitope.png
Figure 10. Beta chains (first column) and respective α chains (fifth column) from 3 ME/CFS patients (L4-02, L4-10, and L3-20, last column).

So, we have seen so far the first two steps of the universal immune testing in ME. What about the third step? In his talk, Mark Davis didn’t present any particular epitope, he just showed a slide with what likely is the selection of the epitopes from the peptide library (see paragraph 6.3) by one of the TCRs reported in figure 10. This selection is reported in figure 11, but from that picture, it is not possible to gather any information about the identity of these epitopes. As you probably remember from paragraph 6.3, the analysis of the peptides selected by a TCR among the peptide library allows the identification of a substitution matrix that can be used to select all the possible epitopes of that specific TCR, from the peptide database. This last crucial step has to be performed yet, or it has been already performed, but Davis has not communicated the preliminary results during his talk. Recently new resources have been made available by Open Medicine Foundation, for this promising research to be further pursued, among other projects (R). The aim here, as already said, is to find the antigen that triggers this T cell response. As Mark Davis said, it might be an antigen from a specific pathogen (perhaps a common pathogen that comes and goes) that elicits an abnormal immune response which ends targeting some host tissue (microglia, for instance), thus leading to the kind of immune activation that has been recently reported by Mark Davis himself and others in ME/CFS (Montoya JG et al. 2017). The idea of a common pathogen triggering a pathologic immune response is not new in medicine, and rheumatic fever (RF) is an example of such a disease: RF is an autoimmune disease that attacks heart, brain and joints and is generally triggered by a streptococcal throat infection (Marijon E et al. 2012). The other possible avenue is, of course, that of an ongoing infection of some kind, that has yet to be detected. As said (see par. 6.1), CD8 T cell clonal expansion is present in both acute infections (like early Lyme disease) and autoimmune diseases (like MS) (figure 4), so we have to wait for the antigen identification if we want to understand if the CTLs activity is against a pathogen and/or against a host tissue.

Figure 11. In this picture, we can see the selection, through several rounds, of a bunch of peptides by a particular TCR from a ME patient. The selection takes place among a huge collection of peptides presented by HLA-A2 (MHC I) expressed by yeast cells. At each round the number of possible peptides is smaller.

8. Chronic Lyme does exist

It has probably been overlooked that in his talk, Mark Davis reported also very interesting data on post-treatment Lyme disease syndrome (PTLDS, also known as chronic Lyme disease). In particular, he found a marked clonal expansion in CD8 T cells of 4 PTLDS patients (about 40% of total CTLs) as reported in figure 12: consider that in this case, blue slices represent unique T cells, while all the other slices represent clones! All that has been said about CD8 clonal expansion in ME/CFS does apply in this case too: it might be the proof of an ongoing infection – perhaps the same B. burgdorferi, as suggested by several animal models (Embers ME et al. 2017), (Embers ME et al. 2012), (Hodzic E et al. 2008), (Yrjänäinen H et al. 2010) – or a coinfection (a virus?) or it could be the expression of an autoimmune reaction triggered by the initial infection. This has still to be discovered, running the complete universal immune testing, but what is already clear from figure 12 is that PTLDS is a real condition, with something really wrong going on within the immune response: chronic Lyme does exist.

Figure 12. CD8 T cells clonal expansion in four chronic Lyme patients: there is a marked clonal expansion that stands for T cell activity against a pathogen or against host tissue.

9. Conclusions

Mark Davis and other researchers have developed a complex assay that is able to sequence TCRs from patients, cluster them into groups of TCRs that react to the same antigens, and discover the antigens that triggered that particular T cell response. This assay is a kind of universal immune testing that is theoretically able to recognize if a person (or a group of patients) presents an immune response against a pathogen or against one of his own tissues (or both). This approach has already given pilot data on an ongoing CD8 T cell activity in ME/CFS patients and in chronic Lyme patients and will hopefully identify the trigger of this immune response in the near future. Whether ME/CFS is an ongoing infection, an autoimmune disease or both, the universal immune testing might be able to tell us. This new technology is for immunology, what whole genome sequencing is for genetics, or metabolomics is for molecular diseases: it doesn’t search for a particular pathogen or a particular autoimmune disease. No, it searches for all possible infections and immune disorders, even those that have yet to be discovered.


Consider supporting this website with a donation.


Esame del liquor nella malattia di Lyme

Esame del liquor nella malattia di Lyme

Sierologia su liquido cerebrospinale

Nei pazienti in cui si sospetti la possibilità che il batterio abbia raggiunto il sistema nervoso centrale, si può ricorrere all’analisi del liquido cerebrospinale (CSF). Infatti, la rilevazione di una produzione intratecale di anticorpi specifici contro B. burgdorferi sl dimostra un coinvolgimento, presente o passato, del sistema nervoso centrale e può fornire la base per la diagnosi di neuroborreliosi (1). La produzione intratecale può essere dimostrata prelevando contemporaneamente il siero e il liquido cerebrospinale e diluendoli finché non presentino la medesima concentrazione di IgG. Quindi si procede alla ricerca di anticorpi specifici per B. burgdorferi sl in entrambi i campioni e infine si calcola il rapporto tra i valori ottenuti su liquor e quelli ottenuti su siero, ricavando l’indice CSF/siero. Se questo indice è >1.3, allora la produzione intratecale è dimostrata (2). Questo tipo di analisi può essere discriminante in alcune circostanze, infatti sono stati descritti casi di Lyme in cui si ha produzione intratecale di anticorpi, in pazienti sieronegativi (3). La rilevazione di IgG intratecali ha una sensibilità dall’80 al 100% nella neuroborreliosi. Nei bambini con neuroborreliosi precoce, la produzione intratecale di IgM specifiche per B. burgdorferi sl ha anch’essa una elevata specificità (4). La produzione di anticorpi intratecali sembra associata con maggiore frequenza alla infezione da B. garinii (5) ed è infatti un riscontro raro negli Stati Uniti. Una falsa positività nella produzione intratecale delle IgM è stata riportata nel caso di meningite da Epstein-Barr virus (6).

Parametri infiammatori del liquido cerebrospinale

Nella neuroborreliosi europea si riscontra tipicamente una elevata conta leucocitaria nel liquido cerebrospinale, dell’ordine di 10-1000 leucociti/ml. L’elevazione dei leucociti è dovuta principalmente a un aumento di linfociti in generale e di cellule B del plasma, in particolare (6). Nella neuroborreliosi si riscontra altresì un aumento delle proteine nel liquido cerebrospinale e la presenza di bande oligoclonali (6), dove entrambi i parametri sono marcatori di infiammazione del sistema nervoso centrale. La presenza di bande oligoclonali nel liquido cerebrospinale indica la produzione intratecale di immunoglobuline della classe IgG, che non sono presenti nel siero. Le bande oligoclonali non sono specifiche della neuroborreliosi di Lyme, essendo presenti anche in altre patologie neuroimmuni. Si consideri infatti che il 90% di pazienti affetti da sclerosi multipla è positiva al test delle bande oligoclonali (7).


Considera una donazione per sostenere questo blog.



  1. Wilske, B, Fingerle, V e Schulte-Spechtel, U. Microbiological and serological diagnosis of Lyme borreliosis. FEMS Immunol Med Microbiol. Feb 2007, Vol. 49, 1, p. 13-21.
  2. Wilske, B, et al. Intrathecal production of specific antibodies against Borrelia burgdorferi in patients with lymphocytic meningoradiculitis (Bannwarth’s syndrome). J Infect Dis. Feb 1986, Vol. 153, 2.
  3. Wilske, B, Zoller, L e Brade, V. MIQ 12 Lyme-borreliose. [aut. libro] H Mauch e R Luticken. Qualitätsstandards in der mikrobiologisch-infektiologischen Diagnostik. Verlag : Munich: Urban&Fischer, 2000.
  4. Hansen, K e Lebech, AM. Lyme neuroborreliosis: a new sensitive diagnostic assay for intrathecal synthesis of Borrelia burgdorferi–specific immunoglobulin G, A, and M. Ann Neurol. 1991, Vol. 30, 2.
  5. Ruzić-Sabljić, E, et al. Analysis of Borrelia burgdorferi sensu lato isolated from cerebrospinal fluid. APMIS. 2001, Vol. 109, 10.
  6. Mygland, A, et al. EFNS guidelines on the diagnosis and management of European Lyme neuroborreliosis. European Journal of Neurology. 2010, Vol. 17, p. 8,16.
  7. Johns, Paul. Clinical Neuroscience. s.l. : Elsevier, 2014.


Western blot per malattia di Lyme

Western blot per malattia di Lyme
  • L’immunoblot (o western blot) è una procedura che consente di determinare contemporaneamente la presenza, nel sangue analizzato, di anticorpi contro più antigeni del patogeno. Una rassegna di questi antigeni si trova qui. Gli antigeni sono disposti su un supporto, divisi in fasce trasversali (dette bande) contigue ma separate, mentre l’esecuzione del test non è dissimile da quella descritta per il test ELISA. In Figura 1 e in Figura 2 sono rappresentati i supporti (IgM e IgG, rispettivamente) per un western blot attualmente in commercio in Europa (Euroimmun).
Figura 1. Rappresentazione di un western blot europeo per IgM. Disegno di Paolo Maccallini.
Figura 2. Rappresentazione di un western blot europeo per IgG. Disegno di Paolo Maccallini.

Affinché un western blot (in IgM o in IgG) sia considerato positivo, non è necessario che tutte le bande lo siano, sono sufficienti solo alcune bande positive. Il numero e il tipo di bande positive affinché un western blot sia positivo è stato determinato con considerazioni statistiche in modo da garantire la massima specificità del test, determinando dei criteri validi per l’area europea. Tali criteri sono riportati in Tabella 1, dove sono posti a confronto con i criteri adottati negli Stati Uniti (CDC, 1995). Come si vede i criteri europei richiedono la positività a un minor numero di bande, essendo la risposta immunologica dei pazienti affetti da borrelia europea ristretta a un numero inferiore di proteine di superficie (Wilske B et al. 2007). In Tabella 2 e Tabella 3 sono riportate sensibilità e specificità per un western blot tedesco attualmente in uso, interpretato con criteri europei (Branda JA et al. 2013). Nella prima tabella le IgM e le IgG sono considerate separatamente, mentre nella seconda si considera un test positivo se positivo in almeno una delle due classi di anticorpi. Dalla Tabella 3 si può notare come, ad esempio, un western blot negativo sia un falso negativo nel 13% di casi neuroborreliosi, così come un test positivo sarà un falso positivo nel 9% di casi di neuroborreliosi o artrite di Lyme.

criteri IB
Tabella 1. Criteri per l’interpretazione del western blot in Europa e in USA. Tabella di Paolo Maccallini.
sensibilità specificità 1
Tabella 2. Sono riportati sensibilità e specificità per un western blot europeo. EM sta per eritema migrante, NB per neuroborreliosi, ACA per acrodermatite cronica atrofica. IgG e IgM sono considerate separatamente. Tabella di Paolo Maccallini.
sensibilità specificità 2
Tabella 3. Sono riportati sensibilità e specificità per un western blot europeo. EM sta per eritema migrante, NB per neuroborreliosi, ACA per acrodermatite cronica atrofica. IgG e IgM sono considerate insieme. Tabella di Paolo Maccallini.

Risonanza magnetica spettroscopica e neuroinfiammazione

Risonanza magnetica spettroscopica e neuroinfiammazione

Equazione di Larmor

Quando un tessuto (ad esempio il cervello) è inserito in un campo magnetico esterno (che indichiamo B), il momento angolare di spin dei nuclei degli atomi che lo compongono descrive un moto di precessione attorno alla direzione di B, con una frequenza angolare Ω data dalla equazione di Larmor:

Ω = ϒ×B

dove ϒ (detta rapporto giromagnetico) è una caratteristica del nucleo. Ad esempio, il nucleo di idrogeno ha una ϒ di 267.513×106 rad s−1 T −1, il nucleo di carbonio ha una ϒ di 67.262×106 rad s−1 T −1, etc.

Risonanza magnetica

Quando un nucleo inserito in un campo magnetico esterno B viene esposto a un radiazione elettromagnetica con frequenza pari alla sua frequenza angolare omega (indicata dalla equazione di Larmor), allora il nucleo produrrà a sua volta una emissione elettromagetica. Negli apparecchi di risonanza magnetica usati in ospedale, il segnale emesso dai nuclei di idrogeno presenti nei tessuti inseriti in un campo magnetico esterno, viene intercettato da antenne ed elaborato da un software che permette di localizzarne la posizione relativa, ricostruendo una immagine del tessuto.

Risonanza magnetica spettroscopica

Come detto, le emissioni elettromagnetiche dei nuclei di idrogeno dei tessuti sottoposti a un campo magnetico esterno fisso e a una emissione elettromagnetica avente una frequenza pari alla loro frequenza angolare omega, sono usate per produrre una immagine del tessuto stesso. Tuttavia è possibile utilizzarle anche per fornire una indicazione sulla concentrazione di alcune specie chimiche in esso contenute. Questo è fattibile grazie al fenomeno detto ‘spostamento chimico’ (chemical shift) il quale è legato alla interferenza che il contesto molecolare del nucleo atomico esercita sul campo magnetico in cui esso è inserito. Senza entrare nei dettagli, si consideri che a seconda della molecola a cui il nucleo di idrogeno appartiene, il campo magnetico esterno sarà alterato, e dunque risulterà alterata anche la frequenza angolare fornita dalla equazione di Larmor. A seconda del valore assunto da omega sarà allora possibile capire se gli atomi di idrogeno appartengono a molecole di colina, o di lattato etc. Su questo principio si basa la risonanza magnetica spettroscopica, che consente di fornire la concentrazione di varie specie chimiche in una regione (in genere un centimetro cubo) di tessuto, che sia un cervello o un muscolo scheletrico. In figura 2 abbiamo ad esempio la concentrazione di diverse specie chimiche in un centimetro cubo di materia bianca del cervello di un essere umano (Bertholdo, D et al.), (Safriel Y et al. 2005).

Figura 2. Risonanza magnetica spettroscopica di un cervello umano, con indicazione della concentrazione di colina (Cho), creatina (Cr) e lattato (Lac). Echo time (TE) di 135 ms.

Risonanza magnetica spettroscopica e malattie

La risonanza magnetica spettroscopica del cervello può essere utilizzata per la diagnosi di malattie metaboliche, attraverso la misura del lattato. Nella ME/CFS ad esempio è stato possibile dimostrare un incremento significativo del lattato nei ventricoli laterali dei cervelli dei pazienti (Mathew, 2009). Altre applicazioni possibli sono quelle oncologiche, infatti lesioni neoplastiche del cervello sono associate a un aumento di colina (Cho) e a una riduzione di N-acetilaspartato (NAA). In questi casi la risonanza magnetica spettroscopica permette sia una diagnosi precoce, che il monitoraggio della risposta alle terapie (Bertholdo, D et al.).

Risonanza magnetica spettroscopica nella ME/CFS e nella Lyme

Di particolare interesse  per malattie come la ME/CFS e la malattia di Lyme potrebbe essere la correlazione tra un aumento di colina (o meglio, del suo rapporto con la creatina, Cho/Cr) e fenomeni infiammatori. Nella ME/CFS un livello elevato del rapporto Cho/Cr è stati riscontrato sia nella corteccia occipitale (Puri BK et al. 2002) che nei gangli della base (Chaudhuri BR et al. 2003), (Tomoda A et al. 2000). In questi studi non si apprezzavano alterazioni anatomiche concomitanti. Nella neuroboreliosi, uno studio del 2004 su 12 pazienti ha rilevato un incremento significativo del rapporto Cho/Cr nella materia bianca dei lobi frontali, in assenza di lesioni apprezzabili del parenchima (Ustymowicz A et al. 2004). Alterazioni simili sono state descritte nella neurite ottica (Tourbah A et al. 1999) e nella infezione da virus dell’epatite C (Forton DM et al. 2001). Forton e colleghi ipotizzarono che l’origine di queste alterazioni fosse da ricondurre a uno stato di attivazione della microglia. Un aumento del rapporto Cho/Cr in pazienti con HIV è stato attribuito a un aumento di attività delle cellule gliali (astrociti, oligodendrociti, cellule di Schwann e microglia) (Chang L et al. 1999) ed è ritenuto riflettere la presenza di fenomeni infiammatori (Lee PL et al. 2003). Un aumento del rapporto Cho/Cr è stato riscontrato anche in lesioni infiammatorie nella toxoplasmosi cerebrale (Roberto J et al. 2009).

Figura 3. Volume di 64 cm cubici nel la materia bianca sopraventricolare di un maschio con clinica ME/CFS e infezione attiva da Borrelia burdorferi. TE: 135 ms, campo magnetico di 3 T. Si nota una regione dell’emisfero destro con aumento di Cho/Cr, probabilmente riconducibile a una lesione infiammatoria.

Il caso di un paziente

In figura 3 abbiamo il caso di un uomo con clinica compatibile con ME/CFS e infezione attiva da Borrelia burgdorferi (PCR positiva su sangue periferico). L’immagine è stata ottenuta con un campo magnetico di 3 tesla e un echo time di 135 ms. Si può notare nell’emisferso destro (a sinistra dell’osservatore) una regione della materia bianca sopraventricolare in cui il valore del rapporto Cho/Cr cresce rispetto al tessuto circostante e rispetto alla medesima regione dell’emisfero sinistro. Il valore risulta elevato anche rispetto al valore medio del Cho/Cr di controlli sani per la stessa sede anatomica (centrum semiovale), che è di 1.05, con una deviazone standard di 0.13 (Lee PL et al. 2003).  In assenza di lesioni macroscopiche visibili e di qualunque sospetto di neoplasia, questa alterazione potrebbe essere riconducibile a un processo infiammatorio legato a una infezione cerebrale da Borrelia burgdorferi, il cui DNA è stato rilevato nel sangue del paziente pochi mesi dopo questo esame. Si osserva che il paziente riferiva episodi di irrigidimento dell’emisoma controlaterale (il sinistro). Il suo sistema immunitario innato è risultato profondamente alterato con livelli di IL-1beta, IL-8, IL-6, MCP1, MIP-1beta che sono da decine a centinaia di volte oltre il massimo del valore normale. Il suo sistema immunitario adattivo produce anticorpi contro il sistema nervoso centrale.

La conta delle NK CD57+ nelle malattie autoimmuni, nella ME/CFS e nella malattia di Lyme

La conta delle NK CD57+ nelle malattie autoimmuni, nella ME/CFS e nella malattia di Lyme


Alcuni laboratori offrono la conta delle cellule natural killer che esprimono il CD57 (NK CD57+ o anche CD3-CD57+) come test di supporto alla diagnosi della malattia di Lyme e, soprattutto, come marcatore utile a monitorare la risposta del paziente alla terapia antibiotica. Valori bassi della conta assoluta di questa sottopopolazione linfocitaria sarebbero indicativi di una infezione in corso da B. burgdorferi sl. Vediamo brevemente quali sono i dati a favore e quelli a sfavore dell’uso di questo esame immunitario nella malattia di Lyme.

Figura 1. Rappresentazione qualitativa delle sottoclassi delle cellule NK, identificate dagli antigeni CD56, CD57 e CD16. Si noti che la sottopopolazione che esprime il CD57 è costituita prevalentemente da NK che esprimono contemporaneamente il CD56DIM e il CD16. Disegno di Paolo Maccallini.

Cosa si intende per CD57?

Il CD57 è una glicoproteina rinvenuta per la prima volta nel 1981 sulla superficie delle cellule natural killer (NK) (Abo, T et Balch CM, 1981), un tipo di globuli bianchi coinvolti principalmente nella difesa contro infezioni intracellulari. Il CD57 è presente solo in una sottopopolazione di NK (Figura 1), ma è espresso anche da alcune cellule T CD4+ (Velardi A et al. 1986) e T CD8+ (Clement LT et al. 1984) (Figura 2), nonché da varie cellule del sistema nervoso centrale e periferico dei vertebrati (Schuller-Petrovic S et al. 1983).

Figura 2. Sottopopolazioni linfocitarie che esprimono il CD57. Si noti che le NK CD57+ possono anche essere designate CD3-CD57+, in quanto non esprimono l’antigene CD3, a differenza delle cellule T CD57+, che invece possono essere designate CD3+CD57+. Disegno di Paolo Maccallini.

NK CD57+ e malattia di Lyme

Nella malattia di Lyme è stato riportata una riduzione della conta delle NK CD57+ (ma non delle cellule T CD8+CD57+ né delle NK CD56) in uno studio su 73 pazienti con sintomi persistenti della malattia di Lyme (Stricker RB et Winger EE 2001) e in almeno due case study (Stricker RB et al. 2002), (Stricker RB et Winger EE 2003). Questi dati tuttavia non sono stati confermati in uno studio successivo su nove pazienti Lyme con sintomi cronici (Marques A et al. 2009). Stricker e colleghi osservarono che i valori più bassi della conta delle NK CD57+ competeva a pazienti Lyme con sintomi cronici di tipo neurologico, e che un trattamento antibiotico in grado di portare i pazienti alla remissione, comportava anche la normalizzazione della conta delle NK CD57+ (vedi Figura 3). Da queste rilevazioni (per ora non confermate da altri autori) e dalla constatazione di valori normali delle NK CD57+ nella infezione acuta da B. burgdorferi sl, Stricker e colleghi hanno proposto la conta delle NK CD57+ come un marcatore specifico e sensibile della persistenza di una infezione cronica nella malattia di Lyme (Stricker RB et Winger EE 2001).

Figura 3. Conta delle NK CD57+ in pazienti Lyme con sintomi cronici di tipo neurologico (LDN), in Lyme con sintomi cronici muscolo-scheletrici (LDM), in pazienti Lyme in remissione (LDR), in controlli adulti sani (HV), in bambini con disordini del spettro autistico (ADS) e in una popolazione pediatrica sana (PHV). I dati dei gruppi LDN, LDM e LDR sono tratti da (Stricker et Winger 2001), quelli del gruppo HV sono da (Marques A et al. 2009). Quelli dei gruppi ASD e PHV sono da (Siniscalco D et al. 2016). I dati sono riportati come media (mean) più/meno una deviazione standard (SD). Grafico di Paolo Maccallini.

Perché le NK CD57+ diminuiscono?

Una possibile giustificazione del motivo per il quale le NK CD57+ dovrebbero ridursi di numero nella infezione cronica da B. burgdorferi sl è stata proposta dallo stesso Stricker, il quale ha notato che questi leucociti sono inibiti dalle citochine sintetizzate dalle T CD4+, ovvero INF-γ, TNF-α e IL-2; dunque una loro riduzione potrebbe riflettere una stimolazione cronica delle T CD4+ da parte di B. burgdorferi sl (Stricker RB et al. 2002). Questa ipotesi è consistente con l’osservazione che nella malattia di Lyme si manifesta una attivazione selettiva delle cellule T CD4+ (Yssel H et al. 1991). De Meirleir ha proposto invece uno scenario in cui la prostaglandina E2 sopprime la risposta imunitaria delle NK, tra cui in particolare quella delle NK CD57+ (De Meirleir K et al. 2016, in IACFS/ME 2016 syllabus, pag 30).

NK CD57+ in altre patologie

Sebbene sia stato riportato dagli stessi autori che tale conta risulta normale in diverse patologie infettive e autoimmuni, è tuttavia da rilevare che recentemente un ridotto numero di NK CD57+ è stato riportato anche in bambini con manifestazioni dello spettro autistico (Siniscalco D et al. 2016) (vedi Figura 3). Inoltre una deplezione dei leucociti che esprimono il CD57 è stata in passato riferita in malattie autoimmuni come il lupus eritematoso sistemico e la sindrome di Sjögren (Struyf NG et al. 1990) (vedi figura 4). In base a quest’ultima osservazione, una misurazione contemporanea delle NK CD57+ e delle cellule T CD57+ sembrerebbe dunque necessaria a discriminare tra malattie autoimmuni e infezione cronica da B. burgdorferi sl. Recentemente De Meirleir ha proposto uno scenario in cui la prostraglandina E2 indotta da lipopolisaccaridi provenienti da batteri intestinali deprime le cellule NK, e in particolare riduce il numero delle NK CD57+, nei pazienti affetti da ME/CFS (De Meirleir K et al. 2016, in IACFS/ME 2016 syllabus, pag 30). Nella ME/CFS, due studi hanno dimostrato una riduzione del numero delle cellule Natural Killer che esprimono il CD57 (Gupta S et Vayuvegula B 1991), (Tirelli U et al. 1994).

Figura 4. Conta delle CD45+ CD57+ nella syndrome di Sjögren (SS), nel lupus eritematoso sistemico (SLE) e nei controlli sani (HV). I dati sono presi da (Struyf NG et al. 1990) e sono stati convertiti da media + errore standard a media (mean) più/meno deviazione standard (SD), per renderli omogenei a quelli della figura 3, anche se qui non abbiamo una distribuzione gaussiana.
Figura 5. Sottopopolazioni linfocitarie di una paziente con clinica CFS e PCR positiva per Borrelia burgdorferi. Le CD45+ sono complessivamente ridotte. I valori dele CD3-CD57+ sono stati ottenuti sottraendo il valore delle CD3- CD57+ a quello delle CD45+ CD57+. Tabella di Paolo Maccallini.

Discussione di un caso

Come abbiamo visto la riduzione del numero delle NK CD57+ non è un marcatore specifico della malattia di Lyme, infatti l’insieme dei linfociti che esprimono il CD57+ (ovvero i CD45+CD57+) risultano complessivamente depressi in almeno due patolgie autoimmuni (Lupus sistemico e Sjögren), mentre le NK CD57+ sono depresse nei disturbi dello spettro autistico (dove però non è stata effettuata una misura dei CD45+CD57+). Un modo per aumentare la specificità di questo marker sarebbe quello di misurare contemporaneamente le NK CD57+ e le T CD57+ (oppure le CD45+CD57+), infatti Stricker e colleghi dimostrarono che nella Lyme cronica le prime risultavano depresse, mentre le seconde non lo erano, e questa sembra una circostanza peculiare della Lyme (che andrebbe però confermata da studi indipendenti). In figura 5 riporto le sottopopolazioni linfocitarie di una paziente con clinica ME/CFS e PCR positiva per Borrelia burgdorferi. Come si vede le sue NK CD57+ sono molto basse (21,2 cellule/microlitro). Tuttavia si vede anche che tutti i suoi linfociti che esprimono il CD 57 sono depressi (105 cellule/microlitro), mentre le cellule T che esprimono il CD57 (le CD3+CD57+) sono nella norma. Quindi il presente caso  rientra sia nel profilo descritto da Stricker e colleghi nel 2001 (Stricker et Winger 2001), che in quello riscontrato nelle malattie autoimmuni (Struyf NG et al. 1990).


La conta delle NK CD57+ non costituisce un marcatore specifico della malattia di Lyme, infatti tutti i linfociti che esprimono il CD57 sono depressi anche nelle malattie autoimmuni. Un modo per aumentare la specificità di questo marcatore è quello di misurare contemporaneamente anche il livello delle CD3+CD57+, poiché nella malattia di Lyme è stata riportato un decremento delle NK CD57+, ma non delle CD3+CD57+.



Considera una donazione per sostenere questo blog.