j.mehy.2017.10.024

Accepted Manuscript
Autoimmunity against a glycolytic enzyme as a possible cause for persistent
symptoms in Lyme disease
Paolo Maccallini, Serena Bonin, Giusto Trevisan
PII:
DOI:
Reference:
S0306-9877(17)30771-5
https://doi.org/10.1016/j.mehy.2017.10.024
YMEHY 8719
To appear in:
Medical Hypotheses
Received Date:
Accepted Date:
20 July 2017
24 October 2017
Please cite this article as: P. Maccallini, S. Bonin, G. Trevisan, Autoimmunity against a glycolytic enzyme as a
possible cause for persistent symptoms in Lyme disease, Medical Hypotheses (2017), doi: https://doi.org/10.1016/
j.mehy.2017.10.024
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Autoimmunity against a glycolytic enzyme as a possible cause for persistent symptoms in
Lyme disease
Paolo Maccallini1, Serena Bonin2 and Giusto Trevisan2
1: Department of Mechanical Engineering, Sapienza University of Rome-Rome-Italy
2: DSM-Department of Medical Sciences-Unit of Dermatology-University of Trieste-Trieste-Italy
Corresponding author:
Serena Bonin
University of Trieste
DSM-Department of Medical Sciences
Cattinara Hospital
Surgical Pathology BLG
Strada di Fiume 447
34149 Trieste
Italy
Abstract
Some patients with a history of Borrelia burgdorferi infection develop a chronic symptomatology
characterized by cognitive deficits, fatigue, and pain, despite antibiotic treatment. The pathogenic
mechanism that underlines this condition, referred to as post-treatment Lyme disease syndrome
(PTLDS), is currently unknown. A debate exists about whether PTLDS is due to persistent infection
or to post-infectious damages in the immune system and the nervous system. We present the case of
a patient with evidence of exposure to Borrelia burgdorferi sl and a long history of debilitating
fatigue, cognitive abnormalities and autonomic nervous system issues. The patient had a positive
western blot for anti-basal ganglia antibodies, and the autoantigen has been identified as γ enolase,
the neuron-specific isoenzyme of the glycolytic enzyme enolase. Assuming Borrelia own surface
exposed enolase as the source of this autoantibody, through a mechanism of molecular mimicry,
and given the absence of sera reactivity to α enolase, a bioinformatical analysis was carried out to
identify a possible cross-reactive conformational B cell epitope, shared by Borrelia enolase and γ
enolase, but not by α enolase. Taken that evidence, we hypothesize that this autoantibody interferes
with glycolysis in neuronal cells, as the physiological basis for chronic symptoms in at least some
cases of PTLDS. Studies investigating on the anti-γ enolase and anti-Borrelia enolase antibodies in
PTLDS are needed to confirm our hypotheses.
Keywords: Post-treatment Lyme disease syndrome; autoimmunity; molecular mimicry; enolase;
glycolysis.
Introduction
Borrelia burgdorferi (Bb), the causative agent of Lyme disease (LD), infects humans through
Ixodes tick bites. Infection can involve several tissues, including skin, joints, heart and nervous
system and can result in arthritis, carditis, and neurological symptoms (1). Involvement of nervous
system, known as neuroborreliosis, may manifest as encephalopathy, myelopathy, and peripheral
neuropathy, which usually respond to antibiotic treatment (2). Nevertheless, some patients with LD
report only partial and transient improvements in cognitive impairment, fatigue and musculoskeletal
pain with prescribed courses of antibiotics and develop a chronic condition referred to as posttreatment Lyme disease syndrome (PTLDS) (3). It is currently under debate if a partial response to
antimicrobials could be due to persistent infection, damage to the brain, or to some form of
1
acquired immune dysfunction (4). In detail, several studies have supported the hypothesis that some
sort of immune dysregulation could occur. Both T and B cell autoimmunity have been detected in
drug-resistant Lyme arthritis (5, 6) and several authors have reported on the potential of
autoimmunity against neural antigens due to molecular mimicry, in LD: cross-reactivity has been
experimentally shown between B. burgdorferi’s flagellin and a component of human peripheral
nerve axon (heat shock protein 60) (7), between non-protein antigens of the pathogen and
gangliosides (8), and between Borrelia OspA and antigens expressed in human brain, spinal cord
and dorsal root ganglia (9). Recently, two studies reported heightened, but apparently non-specific,
production of antibodies against brain proteins in about 50% of patients with a history of LD and
persistent symptoms (10, 11). Here, we report the detection of anti-γ enolase antibodies in one
patient with evidence of Borrelia burgdorferi infection and persistent symptoms of fatigue and
cognitive impairment. Mammalian enolase acts as homodimeric or heterodimeric isoenzymes,
combining three subunits: α, β, and γ enolase. Alpha-enolase is ubiquitous, β enolase is mainly
expressed in muscles, while γ enolase in neurons and neuroendocrine tissues (12). In absence of
serum reactivity against the α subunit of the same enzyme, and assuming the enolase of B.
burgdorferi as the trigger for the autoantibody, a bioinformatical analysis with an ad-hoc-developed
software has been carried out to propose a cross-reactive conformational epitope between γ enolase
and Borrelia burgdorferi enolase and to suggest a possible pathogenic mechanism for chronic
symptoms in Lyme disease.
Materials and methods
Patient
A man (aged 35 at diagnosis), with a long-lasting history of fatigue and cognitive issues, presented
at the University Hospital of Trieste, where he was submitted to a complete clinical and
instrumental evaluation. Infectious agents were searched for in the patient’s serum as shown in
Table 1. DNA from peripheral blood was submitted to Borrelia detection as previously reported
(13). Anti-Borrelia burgdorferi antibodies were detected by Western blot using anti-BorreliaEuroline-RN-AT (Euroimmun, Germany).
Anti-basal ganglia antibodies (ABGA) assay was performed as previously described (14). Antigens
were derived from sections of human basal ganglia containing caudate and putamen; the sample
comes from a donor with no history or evidence of neurological disease. Serum was also analysed
for reactivity to α enolase, γ enolase, and pyruvate kinase, which have been linked respectively to
the 45 kDa (doublet) and the 60 kDa bands of the ABGA immunoblot (15). Those tests were
performed at the Carlo Besta Institute (Neuromuscular and Neuroimmunology Laboratory), except
for the anti-α enolase assay, that was performed by Wieslab Laboratory, Sweden. An extensive
panel of other autoantibodies (including various anti-neuronal Abs) was also performed (Table 2).
Search for unique peptides of γ enolase
A software, called EPITOPE, written by one of the authors (MP) in Octave 4.0.0, has been used to
identify a set of linear peptides of γ enolase, which are not shared with α enolase. EPITOPE
considers each possible peptide of γ enolase with a length of δ amino acids, and calculates the score
of the best alignment with every α enolase peptides of the same length. For each comparison,
EPITOPE implements the algorithm developed by Needleman and Wunsch (16), with a gap model
ܽ + ܾ ⋅ ‫ݔ‬, where ܽ is the opening gap penalty, ܾ is the extending one, and ‫ ݔ‬is the extension of the
gap. A penalty for gaps at the end of the alignment was also assumed. As a result, a score is
associated to every γ enolase δ-amino-acid peptide. The program then gives as output only those
peptides of γ enolase to which a score below S is associated. We regard those peptides of γ enolase
as non-cross-reactive with α enolase. Parameters have been chosen according to what follows. As α
2
enolase and γ enolase share about 83% of identity (Figure 1), a possible choice for the substitution
matrix is MDM20 (17). Given two random sequences of m and n amino acids respectively, it has
been calculated (18) that the probability p that a local alignment attains a score of at least S is given
by
Eq. 1 ‫ = ݌‬1 − ݁ ି௄⋅௠⋅௡⋅௘
షഊೄ
To define the parameters λ, K, a, and b, LALIGN provided by the European Bioinformatic Institute
was used. If γ enolase is selected as the first sequence, α enolase as the second sequence, MDM20
as substitution matrix, the program gives λ = 0.2298 and K = 0.3234, and selects an opening gap
and an extending gap of 22 and 4, respectively. Alignments were considered significant when
‫ < ݌‬0.05 (17), corresponding to a score ܵ > 60, according to Eq. 1. We calculated δ as the lowest
number of amino acids necessary to give a significant score (minimal significant length), which is
obtained dividing the minimal significant score S by the relative entropy of the substitution matrix
of choice (17). We obtain for δ a value of 7 amino acids.
Running EPITOPE, 162 peptides with a score below 60 have been identified. These peptides cluster
into 8 subsequences of amino acids of γ enolase which include all the three regions of sequence
difference between the human α and γ forms (271-285, 298-316, 416-433), experimentally
demonstrated by McAleese et al. to be immunogens that can be used to generate antibodies to γ
enolase not reacting with α enolase (19). A threshold score of 50 leads to a narrower selection for
unique peptides of γ enolase, which still includes the 3 sequences found by McAleese. Therefore, a
score < 50 corresponds to non-cross-reactive peptides, a score > 60 corresponds to cross-reactive
peptides, while a score 50 ≤ ܵ ≤ 60 defines borderline results. We obtain 93 peptides scoring < 50
that cluster into twelve subsequences of amino acids of γ enolase (Figure 2, columns 1, 2). Using
the experimentally determined 3D structure of γ enolase (20), all twelve subsequences were verified
to be displayed on its surface, in agreement with the fact that sites in the protein cores are highly
evolutionary conserved (21).
Search for cross-reactive epitopes between enolases
A second type of software, EPITOPE-LOCAL, has been written to calculate the best local
alignment between each peptide of γ enolase found by EPITOPE and the following enolases: human
β enolase, enolase from B. afzelii (strain PKo), enolase from B. burgdorferi ss (strain ZS7) and
enolase from B. garinii (strain PBr) (Figure 2, columns 3-6). EPITOPE-LOCAL is a simpler
version of EPITOPE, which aligns a given peptide of δ amino acids with all possible peptides of the
same length from a given sequence, displaying as output the alignment with the highest score. As
EPITOPE, this software is based on the algorithm by Needleman and Wunsch and uses the same
settings (substitution matrix, gap penalties, value for δ). This analysis aimed to find epitopes of
Borrelia enolase that might be a trigger for anti-γ enolase autoantibodies, through a mechanism of
molecular mimicry.
Protein sequences and 3D structures
The following sequences have been used: human α enolase 1 (P06733-1), human β enolase 1
(P13929-1), human γ enolase 1 (P09104-1), Borrelia afzelii (strain PKo) enolase (Q0SNH5-1),
Borrelia burgdorferi ss (strain ZS7) enolase (B7J1R2-1), Borrelia garinii (strain PBr) enolase
(B7XT07). A three-dimensional structure of γ enolase experimentally determined by Leonard P.G.
and colleagues (20) has been considered, the PDB ID of which is 4ZA0. For β enolase, the
experimentally determined structure by Vollmar M. and colleagues (22) (PDB ID: 2XSX) has been
used. Structures for enolases of Borrelia species have not yet been experimentally determined.
Therefore, a theoretical model for B. burgdorferi ss enolase offered by ModBase, which uses as
3
template the experimental 3D structure of enolase from Enterecoccus hirae (PDB ID: 1IYX), has
been considered. To verify the surface exposure of candidate epitopes on α, γ and β enolases,
downloaded in ASN1 format, the 3D viewer Cn3D 4.3.1 has been used.
Results
Patient
The patient is a 37-year-old man. Since 1999 he has been afflicted by severe fatigue, memory
impairment and lack of concentration. At that time, he was 19 and he had never had any
neurological symptom before. In the early years, the abovementioned symptoms presented with a
cyclic trend of relapses and recoveries. In 2002, a flu-like illness precipitated his condition which
has since become chronic. During the last 15 years, he has been experiencing other symptoms, such
as muscular and joint pain, acral hypoesthesia, post-exertional malaise, and orthostatic intolerance,
resulting in him being house-bound. He fulfils the 2015 diagnostic criteria published by IOM (23)
for myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). In August 2014, the patient
presented at Trieste University Hospital, where he resulted positive to Borrelia both by PCR and
Western Blot, as shown in Table 1. The patient was also positive to HHV6 (past infection). Current
infection by Helicobacter pylori was detected and treated (Table 1). In 2000, the patient presented
with a nodule of one centimeter in diameter, located on one nipple. It was investigated with
echography and mammography and cancer was ruled out. Its etiology was not further investigated
at that time and the lesion resolved spontaneously within months. That lesion was compatible with a
nodular form (Spiegler-Fendt type) of lymphadenosis benigna cutis (LABC) of areola mammae,
which is a typical second stage manifestation of LD in Europe (24). Patient was submitted to
antibiotics, with complete and rapid resolution of acral hypoesthesia and only partial and temporary
improvements in the other symptoms. In 2015, an extensive panel of anti-neuronal antibodies was
performed (Table 2). Positive anti-Basal Ganglia Antibodies and anti-γ enolase antibody were
detected.
Cross-reactivity between enolases
EPITOPE identifies 93 peptides of γ enolase that do not cross-react with α enolase (Figure 2,
column 1, 2). Six of them cross-react with enolases from the three main pathogenic species of B.
burgdorferi. With further analyses, we have found that, among these 6, only peptides 49-55, 59-65,
60-66 and 188-194 cross-react with sequences of Borrelia enolases that, in turn, do not cross-react
with any peptide of α enolase (as detailed in Figure 3 for peptide 49-55). The results of this analysis
are summarized in Figure 4. As shown, β enolase is involved in cross-reactivity too, although to a
lesser extent.
Conformational epitopes
Peptides 49-55, 59-65 and 60-66 of one molecule of γ enolase identify a conformational epitope
with peptide 188-194 of the other subunit of the homodimer γ-γ (see Figure 5, left). Peptides 49-55,
59-65 and 60-66 define a conformational epitope on the surface of β enolase, too (not shown).
Peptides 52-58, 62-68 and 63-69 from one molecule of Borrelia burgdorferi enolase generate a
conformational epitope with peptide 186-192 from another molecule, when the two subunits
combine to form the glycolytic dimeric enzyme (Figure 5, right).
If we assume the presence of serum reactivity to both γ enolase and α enolase (which is not the case
of our patient) we find, with the same methods described above, a conformational B cell epitope
shared by all the three human enolases and Borrelia enolase (Figure 6). It partially overlaps with the
previously discussed conformational epitope.
4
Discussion
There is currently debate on whether Lyme patients with long-term symptoms are suffering from
persistent infection and/or from some sort of immune-mediated disorder. These symptoms most
commonly include fatigue, cognitive deficits and musculoskeletal pain (3). Persistent cognitive
abnormalities have been associated to hypometabolism in several brain regions (25, 26). In this
study, we hypothesize that chronic symptoms are caused by an autoantibody - triggered by Borrelia
infection through a mechanism of molecular mimicry - which interferes with glycolysis in neuronal
tissues.
The patient presented in this study had a complex picture of chronic symptoms, which matches the
clinical picture of ME/CFS (23). He had both positive PCR for borrelial DNA and serum antibodies
against two antigens of B. burgdorferi (p41 and p19, Table 1), and responded only partially to
antibiotic treatment. Taken higher level and frequency of anti-neural antibodies in patients with
PTLDS (10, 11), we further investigated for some possible autoimmune mechanism. Reactivity of
patient’s serum against a wide range of neuronal antigens was tested, resulting in positivity to anti-γ
enolase Abs (Table 2). This anti-neural autoimmunity could be due to several factors, but the
observation that enolase is highly conserved from archaebacteria to mammals (27) strongly supports
for an infectious agent as a possible source of the autoantibody, through a cross reactive immune
response between microbial and human enolase. Borrelia enolase (BEN) shares almost 50% of
sequence identity with human enolase and more than 130 similar positions (on a total of 433
residues) (Figure 1). Furthermore, BEN has recently been observed on the outer membrane of the
spirochete (28-30), and has been demonstrated to be immunogenic, both in mice and in humans (30,
31). Therefore, we hypothesize that surface-exposed BEN could elicit antibodies that cross-react
with human γ enolase, due to the similarity between amino acid sequences (molecular mimicry). To
further support our hypothesis, an anti-human α enolase antibody has been shown to recognize
purified recombinant BEN (28).
Our computational analyses allowed us to find a conformational B cell epitope shared by human γ
and Borrelia enolase (and to a lesser extent by β enolase), but not by α enolase (Figure 5). B cell
epitopes, as it is well known, need surface exposure (32). Most of them are conformational and
have linear peptides of maximum 4-7 residues (33). These characteristics are fulfilled by our
conformational epitope. Human enolases are functionally active as five dimers (α-α, α-β, β-β, α-γ,
γ-γ) (12). As the γ-γ dimer is specific for neuronal and neuroendocrine tissues (12), we hypothesize
the central nervous system (CNS) as a possible target for this autoantibody.
Autoantibodies against enolase have already been reported in several autoimmune diseases, while
being relatively rare in healthy controls (15, 34, 35), in neurologic diseases (36), and in autism
spectrum disorder (37). The involvement of enolase autoantibodies on our patient’s symptoms is
supported by the detection of anti-enolase autoimmunity in brain disorders such as PANDAS (15,
38), obsessive-compulsive disorder (OCD) (35, 39), encephalitis lethargica (EL) (34), multiple
sclerosis (MS) (40), and Hashimoto’s encephalopathy (HE) (41-43). In particular, HE resembles
some features of Lyme encephalopathy (25). To support our hypothesis on an infectious
involvement in anti-enolase antibodies production, other authors reported on a causal link between
PANDAS and Streptococcus pyogenes (15, 36, 38). A similar association was proposed for OCD in
adults (35), and for EL (34). Cross-reactivity between streptococcal surface enolase and human α
enolase was experimentally demonstrated in sera of patients with acute rheumatic fever reacting
with both molecules (44).
Although anti-enolase autoimmunity seems to be diffused in a wide range of autoimmune diseases,
different diseases appear to be linked to auto-Ab against specific isoforms: while PANDAS seems
to be mainly associated with anti-γ enolase autoimmunity (15), Behçet’s disease (45), systemic
lupus erythematosus, and systemic sclerosis (46), are linked to antibodies against α enolase, but not
against γ. Furthermore, diseases in which an Ab against the same isoenzyme of enolase is present,
5
have often been linked to different epitopes. For instance, sera from patients with cancer-associated
retinopathy bind residues 56-63 of α enolase (47), while peptide 207-238 of α enolase seems to be a
specific autoepitope in endometriosis (48), and in HE patients, most of sera reacts against the amino
terminal of α enolase (residues 1-134) (41, 43). These data support the hypothesis that the auto-Ab
to the conformational epitope of γ enolase here discussed may be specific for PTLDS.
In our hypothesis, the autoantibody against the conformational epitope of Borrelia enolase binds γ
enolase exposed on the surface of neurons and it is then internalized (Figure 7, III). Once in
cytoplasm, the antibody inhibits the catalytic activity of γ enolase in the ninth step of glycolysis,
depleting the intracellular level of ATP (Figure 7, IV). We propose alteration of brain distal
glycolysis by anti-γ enolase antibodies to explain changes observed in Lyme encephalopathy, where
lower cerebral blood flow and metabolic rate in many areas of the brain have been reported, and
suggested as primarily metabolic driven (25, 26). As γ enolase is expressed on the surface of
neurons and neuroendocrine cells (49), the proposed internalization of anti-enolase autoantibodies is
theoretically possible, and it has in fact been described in vitro, in at least one study (50). In the
same paper, inhibition of enolase enzymatic activity by the internalized autoantibody, and a
consequent reduction in cellular ATP, was reported, in agreement with our hypothesis. To further
support the possible pathogenicity of anti-γ enolase antibody, the putative conformational epitope
here proposed is relatively close to the active site cavity of the dimeric enzyme (51).
Pathogenic effects of autoantibodies against surface-exposed epitopes can also be linked to
activation of complement cascades and Fc receptors (52). This leads to the hypothesis that anti-γ
enolase antibodies could induce damage in neuronal tissues, through the activity of the complement
system and of leukocytes such as natural killer cells and macrophages (Figure 7, II).
We acknowledge as limitations in our study that the analysis presented based on peptides
alignments miss all cross-reactive conformational epitopes due exclusively to protein folding.
Furthermore, we recognize that it is possible that anti-γ enolase antibody found in our patient may
be due to cross-reactivity with antigens from other organisms. In previous studies (10, 11),
heightened production of IgGs against brain proteins has been detected in patients exposed to B.
burgdorferi and with persistent symptoms, but a common autoantigen was not found. That
observation could support the hypothesis that anti-enolase autoimmunity could develop only in a
subset of PTLDS patients, perhaps in those who fulfil diagnostic criteria for ME/CFS, who
represent about 13% of cases (53).
Conclusion
In this study, the presence of antibodies against γ enolase was found in a patient with a complex
picture of neurological symptoms and with exposure to B. burgdorferi. We hypothesize that the
infection by B. burgdorferi elicits an antibody to Borrelia enolase that cross-reacts - via molecular
mimicry - with human γ enolase. A possible cross-reactive conformational epitope has been found
and a possible mechanism for the causal role of this antibody in the genesis of cognitive
manifestations has been suggested.
We acknowledge that our hypothesis needs to be validated in case studies investigating on the
presence of anti-γ enolase antibodies in LD patients with persistent symptoms. The co-presence of
anti-BEN antibodies should also be investigated to support the causal role of BEN for anti-enolase
autoimmunity via molecular mimicry in PTLDS.
Acknowledgments
We are grateful to Margarita Bitetti and Giada Da Ros for their support in this study.
References
6
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
G. Trevisan and M. Cinco, Lyme disease. A general survey, Int J Dermatol 29 (1990), pp. 18.
A. Mygland, U. Ljostad, V. Fingerle, et al., EFNS guidelines on the diagnosis and
management of European Lyme neuroborreliosis, Eur J Neurol 17 (2010), pp. 8-16, e11-14.
J.N. Aucott, L.A. Crowder and K.B. Kortte, Development of a foundation for a case
definition of post-treatment Lyme disease syndrome, Int J Infect Dis 17 (2013), pp. e443449.
I.o. Medicine, Critical Needs and Gaps in Understanding Prevention, Amelioration, and
Resolution of Lyme and Other Tick-Borne Diseases: The Short-Term and Long-Term
Outcomes: Workshop Report, The National Academies Press, Washington, DC (2011).
E.E. Drouin, L. Glickstein, W.W. Kwok, G.T. Nepom and A.C. Steere, Human homologues
of a Borrelia T cell epitope associated with antibiotic-refractory Lyme arthritis, Mol
Immunol 45 (2008), pp. 180-189.
E.E. Drouin, R.J. Seward, K. Strle, et al., A novel human autoantigen, endothelial cell
growth factor, is a target of T and B cell responses in patients with Lyme disease, Arthritis
Rheum 65 (2013), pp. 186-196.
L.H. Sigal, S. Williams, B. Soltys and R. Gupta, H9724, a monoclonal antibody to Borrelia
burgdorferi's flagellin, binds to heat shock protein 60 (HSP60) within live neuroblastoma
cells: a potential role for HSP60 in peptide hormone signaling and in an autoimmune
pathogenesis of the neuropathy of Lyme disease, Cell Mol Neurobiol 21 (2001), pp. 477495.
J.C. Garcia-Monco, R.J. Seidman and J.L. Benach, Experimental immunization with
Borrelia burgdorferi induces development of antibodies to gangliosides, Infect Immun 63
(1995), pp. 4130-4137.
A. Alaedini and N. Latov, Antibodies against OspA epitopes of Borrelia burgdorferi crossreact with neural tissue, J Neuroimmunol 159 (2005), pp. 192-195.
A. Chandra, G.P. Wormser, M.S. Klempner, et al., Anti-neural antibody reactivity in
patients with a history of Lyme borreliosis and persistent symptoms, Brain Behav Immun 24
(2010), pp. 1018-1024.
E. Jacek, B.A. Fallon, A. Chandra, M.K. Crow, G.P. Wormser and A. Alaedini, Increased
IFNalpha activity and differential antibody response in patients with a history of Lyme
disease and persistent cognitive deficits, J Neuroimmunol 255 (2013), pp. 85-91.
V. Pancholi, Multifunctional alpha-enolase: its role in diseases, Cell Mol Life Sci 58 (2001),
pp. 902-920.
P. Pauluzzi, S. Bonin, M.A. Gonzalez Inchaurraga, G. Stanta and G. Trevisan, Detection of
spirochaetal DNA simultaneously in skin biopsies, peripheral blood and urine from patients
with erythema migrans, Acta dermato-venereologica 84 (2004), pp. 106-110.
A.J. Church, F. Cardoso, R.C. Dale, A.J. Lees, E.J. Thompson and G. Giovannoni, Antibasal ganglia antibodies in acute and persistent Sydenham's chorea, Neurology 59 (2002),
pp. 227-231.
R.C. Dale, P.M. Candler, A.J. Church, R. Wait, J.M. Pocock and G. Giovannoni, Neuronal
surface glycolytic enzymes are autoantigen targets in post-streptococcal autoimmune CNS
disease, J Neuroimmunol 172 (2006), pp. 187-197.
S.B. Needleman and C.D. Wunsch, A general method applicable to the search for
similarities in the amino acid sequence of two proteins, J Mol Biol 48 (1970), pp. 443-453.
S.F. Altschul, Amino acid substitution matrices from an information theoretic perspective, J
Mol Biol 219 (1991), pp. 555-565.
S. Karlin and S.F. Altschul, Methods for assessing the statistical significance of molecular
sequence features by using general scoring schemes, Proc Natl Acad Sci U S A 87 (1990),
pp. 2264-2268.
7
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
S.M. McAleese, B. Dunbar, J.E. Fothergill, L.J. Hinks and I.N. Day, Complete amino acid
sequence of the neurone-specific gamma isozyme of enolase (NSE) from human brain and
comparison with the non-neuronal alpha form (NNE), Eur J Biochem 178 (1988), pp. 413417.
P.G. Leonard, N. Satani, D. Maxwell, et al., SF2312 is a natural phosphonate inhibitor of
enolase, Nat Chem Biol 12 (2016), pp. 1053-1058.
J. Echave, S.J. Spielman and C.O. Wilke, Causes of evolutionary rate variation among
protein sites, Nat Rev Genet 17 (2016), pp. 109-121.
M. Vollmar, Krysztofinska, E., Chaikuad, A., Krojer, T., Cocking, R., Von Delft, F.,
Bountra, C., Arrowsmith, C.H., Weigelt, J., Edwards, A., Yue, W.W., Oppermann, U.,
Crystal Structure of Human Beta Enolase Enob, RCSB PDB (To Be Published).
M. Institute of, Beyond Myalgic Encephalomyelitis/Chronic Fatigue Syndrome: Redefining
an Illness %@ 978-0-309-31689-7, The National Academies Press, Washington, DC (2015).
G. Trevisan, P. Cattonar, C. Nobile, V. Perkan and G. Stinco, Dermatological manifestations
of Lyme borrelosis, Acta Dermatovenerologica Alpina, Pannonica et Adriatica 5 (1996), pp.
101-107.
B.A. Fallon, R.B. Lipkin, K.M. Corbera, et al., Regional cerebral blood flow and metabolic
rate in persistent Lyme encephalopathy, Arch Gen Psychiatry 66 (2009), pp. 554-563.
A. Newberg, A. Hassan and A. Alavi, Cerebral metabolic changes associated with Lyme
disease, Nucl Med Commun 23 (2002), pp. 773-777.
M. Piast, I. Kustrzeba-Wojcicka, M. Matusiewicz and T. Banas, Molecular evolution of
enolase, Acta Biochim Pol 52 (2005), pp. 507-513.
A.M. Floden, J.A. Watt and C.A. Brissette, Borrelia burgdorferi enolase is a surfaceexposed plasminogen binding protein, PLoS One 6 (2011), p. e27502.
S.V. Nogueira, A.A. Smith, J.H. Qin and U. Pal, A surface enolase participates in Borrelia
burgdorferi-plasminogen interaction and contributes to pathogen survival within feeding
ticks, Infect Immun 80 (2012), pp. 82-90.
A. Toledo, J.L. Coleman, C.J. Kuhlow, J.T. Crowley and J.L. Benach, The enolase of
Borrelia burgdorferi is a plasminogen receptor released in outer membrane vesicles, Infect
Immun 80 (2012), pp. 359-368.
A.G. Barbour, A. Jasinskas, M.A. Kayala, et al., A genome-wide proteome array reveals a
limited set of immunogens in natural infections of humans and white-footed mice with
Borrelia burgdorferi, Infect Immun 76 (2008), pp. 3374-3389.
J.M. Thornton, M.S. Edwards, W.R. Taylor and D.J. Barlow, Location of 'continuous'
antigenic determinants in the protruding regions of proteins, EMBO J 5 (1986), pp. 409-413.
P. Haste Andersen, M. Nielsen and O. Lund, Prediction of residues in discontinuous B-cell
epitopes using protein 3D structures, Protein Sci 15 (2006), pp. 2558-2567.
R.C. Dale, A.J. Church, R.A. Surtees, et al., Encephalitis lethargica syndrome: 20 new cases
and evidence of basal ganglia autoimmunity, Brain 127 (2004), pp. 21-33.
T.R. Nicholson, S. Ferdinando, R.B. Krishnaiah, et al., Prevalence of anti-basal ganglia
antibodies in adult obsessive-compulsive disorder: cross-sectional study, Br J Psychiatry
200 (2012), pp. 381-386.
A.J. Church, R.C. Dale and G. Giovannoni, Anti-basal ganglia antibodies: a possible
diagnostic utility in idiopathic movement disorders?, Arch Dis Child 89 (2004), pp. 611614.
S. Kalra, P.D. Burbelo, A. Bayat, et al., No Evidence of Antibodies against GAD65 and
Other Specific Antigens in Children with Autism, BBA Clin 4 (2015), pp. 81-84.
H. Nicolini, Y. Lopez, A.D. Genis-Mendoza, et al., Detection of anti-streptococcal,
antienolase, and anti-neural antibodies in subjects with early-onset psychiatric disorders,
Actas Esp Psiquiatr 43 (2015), pp. 35-41.
8
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
R.C. Dale, I. Heyman, G. Giovannoni and A.W. Church, Incidence of anti-brain antibodies
in children with obsessive-compulsive disorder, Br J Psychiatry 187 (2005), pp. 314-319.
F. Forooghian, G. Adamus, M. Sproule, C. Westall and P. O'Connor, Enolase autoantibodies
and retinal function in multiple sclerosis patients, Graefes Arch Clin Exp Ophthalmol 245
(2007), pp. 1077-1084.
A. Fujii, M. Yoneda, T. Ito, et al., Autoantibodies against the amino terminal of alphaenolase are a useful diagnostic marker of Hashimoto's encephalopathy, J Neuroimmunol 162
(2005), pp. 130-136.
H. Ochi, I. Horiuchi, N. Araki, et al., Proteomic analysis of human brain identifies alphaenolase as a novel autoantigen in Hashimoto's encephalopathy, FEBS Lett 528 (2002), pp.
197-202.
M. Yoneda, A. Fujii, A. Ito, H. Yokoyama, H. Nakagawa and M. Kuriyama, High
prevalence of serum autoantibodies against the amino terminal of alpha-enolase in
Hashimoto's encephalopathy, J Neuroimmunol 185 (2007), pp. 195-200.
P.A. Fontan, V. Pancholi, M.M. Nociari and V.A. Fischetti, Antibodies to streptococcal
surface enolase react with human alpha-enolase: implications in poststreptococcal sequelae,
J Infect Dis 182 (2000), pp. 1712-1721.
K.H. Lee, H.S. Chung, H.S. Kim, et al., Human alpha-enolase from endothelial cells as a
target antigen of anti-endothelial cell antibody in Behcet's disease, Arthritis Rheum 48
(2003), pp. 2025-2035.
B. Terrier, N. Degand, P. Guilpain, A. Servettaz, L. Guillevin and L. Mouthon, Alphaenolase: a target of antibodies in infectious and autoimmune diseases, Autoimmun Rev 6
(2007), pp. 176-182.
G. Adamus, D. Amundson, G.M. Seigel and M. Machnicki, Anti-enolase-alpha
autoantibodies in cancer-associated retinopathy: epitope mapping and cytotoxicity on retinal
cells, J Autoimmun 11 (1998), pp. 671-677.
M. Walter, H. Berg, F.A. Leidenberger, K.W. Schweppe and W. Northemann, Autoreactive
epitopes within the human alpha-enolase and their recognition by sera from patients with
endometriosis, J Autoimmun 8 (1995), pp. 931-945.
A. Hafner, N. Obermajer and J. Kos, gamma-1-syntrophin mediates trafficking of gammaenolase towards the plasma membrane and enhances its neurotrophic activity, Neurosignals
18 (2010), pp. 246-258.
A. Magrys, T. Anekonda, G. Ren and G. Adamus, The role of anti-alpha-enolase
autoantibodies in pathogenicity of autoimmune-mediated retinopathy, J Clin Immunol 27
(2007), pp. 181-192.
L. Lebioda and B. Stec, Mapping of isozymic differences in enolase, Int J Biol Macromol 13
(1991), pp. 97-100.
B. Diamond, P.T. Huerta, P. Mina-Osorio, C. Kowal and B.T. Volpe, Losing your nerves?
Maybe it's the antibodies, Nat Rev Immunol 9 (2009), pp. 449-456.
D.I. Bujak, A. Weinstein and R.L. Dornbush, Clinical and neurocognitive features of the
post Lyme syndrome, J Rheumatol 23 (1996), pp. 1392-1397.
9
LEGEND TO THE FIGURES
Figure 1. Identity and similar positions shared by Human and Borrelia Enolases. Henα1: human α
enolase 1 (P06733-1); Henβ1: human β enolase 1 (P13929-1); Henγ1: human γ enolase 1 (P091041); Baen: Borrelia afzelii (strain PKo) enolase (Q0SNH5-1); Bben: Borrelia burgdorferi ss (strain
ZS7) enolase (B7J1R2-1); Bgen: Borrelia garinii (strain PBr) enolase (B7XT07). Identities and
similar positions have been calculated by Clustal Omega, with default settings.
Figure 2. Peptides of human γ enolase not shared with human α enolase. In the first column 7
amino acids peptides of γ enolase which do not cross-react with α enolase (score below 50) are
reported. For each epitope of the first column, the best local alignment with human α enolase 1
(P06733-1), human β enolase 1 (P13929-1), human γ enolase 1 (P09104-1), Borrelia afzelii (strain
PKo) enolase (Q0SNH5-1), Borrelia burgdorferi ss (strain ZS7) enolase (B7J1R2-1), Borrelia
garinii (strain PBr) enolase (B7XT07) has been searched. Those local alignments with crossreactivity are green colored, while those which are not cross-reactive are in orange. Borderline
reactivity is in white. For each alignment, p value have been included.
Figure 3. Study of cross-reactivity for epitope 49-55 of γ enolase.
Figure 4. Cross-reactivity among enolases. Cross-reactions between Borrelia enolases and β enolase
are marked in red, cross-reactions between Borrelia enolases and γ enolase are marked in blue.
Cross-reactions between γ and β enolase are marked in black.
Figure 5. The conformational epitope shared by human γ enolase and Borrelia enolase, but not by α
enolase. On the left: peptides 49-55, 59-65, 60-66 and 188-194 of γ enolase form a conformational
epitope (in yellow) when two molecules of γ enolase are bound. On the right: peptides 52-58, 6268, 63-69 and 186-192 of Borrelia burgdorferi enolase form a conformational epitope (in yellow)
when the two subunits form the homodimeric enzyme.
Figure 6. The conformational epitope shared by human enolases and Borrelia enolase. Left:
peptides 37-43, 47-53, and 48-54 of γ enolase form a conformational epitope (in yellow). Center:
peptides 37-43, 47-53, and 48-54 of β enolase form a conformational epitope (in yellow). Right:
peptides 40-46, 50-56, and 51-57 of B. burgdorferi enolase form a conformational epitope (in
yellow). Peptides 37-43, 47-53, and 48-54 identify a similar conformation epitope on α enolase (not
shown).
Figure 7. The proposed pathogenic role of anti-γ enolase antibody in PTLDS. I: Enolase from Bb
binds its specific B cell receptor (BCR), then the antigen is internalized, cut into a T cell epitope
and presented by major histocompatibility complex II (MHC II) to a Th cell with a specific T cell
receptor (TCR). This leads to B cell activation and consequent production of anti-enolase
antibodies. II: The interaction between anti-enolase Ab and surface exposed human enolase activate
complement cascade and leukocytes (e.g. macrophages), through the Fc region, with possible
induction of apoptosis. III: Internalization of enolase-Ab complex. IV: Anti-enolase Ab interferes
with the ninth step of glycolysis.
10
11
12
13
14
15
16
17
Table 1. Infectious diseases workup in this patient (peripheral blood, unless otherwise indicated).
Category
Herpesviruses
Other viruses
Tick borne infections
Pathogen
Test
Herpes simplex virus 1, 2
IgM, IgG
Epstein-Barr virus
IgM, IgG, PCR
Cytomegalovirus
IgM, IgG, PCR
HHV-6
IgM, IgG, PCR
Positive IgG
HHV-7
PCR
Negative
Parvovirus B19
IgM, IgG, PCR
Enterovirus
IgM, IgG
West Nile virus
IgM, IgG
Hepatitis C
Abs
Hepatitis B
Anaplasma
phagocytophilum
Babesia
Abs
PCR
TBE virus
IgM, IgG
Rickettsia spp.
Aspergillus fumigatus
Weil-Felix, PCR
IgM, IgG
PCR
Ab by ELISA
Bartonella henselae
IgM, IgG, PCR
Brucella spp.
IgM, IgG
Campylobacter jejuni
IGA, IgG
Candida albicans
Mannan antigen
Chlamydia pneumoniae
IgM, IgA, IgG
Chlamydia trachomatis
IgM, IgA, IgG
Coxiella burnetii
IgM, IgG
Mycoplasma spp
IgM, IgG, PCR
Streptococcus pyogenes
ASO, throat swab culture
Toxoplasma gondii
IgM, IgG
Treponema pallidum
T.P.H.A.
Helicobacter pylori
stool sample culture
Borrelia burgdorferi sl
Other pathogens
Result
Negative
Negative
Positive (immunization)
IgM, IgG
18
Negative
Positive IgG (p41, p19)
Positive
Negative
Positive
Table 2. Serum autoantibodies workup in this patient.
Category
Autoantibody
Test
Results
Anti-IgG Fc (Rheumatoid factor, RF)
Anti-cyclic citrullinated peptide (CCP)
General autoimmunity
Autoimmune inflammatory
myopathies and scleroderma panel
Anti-nuclear (ANA)
IFI on Hep2
Anti-double stranded DNA (ds-DNA)
EIA
Anti-extractable nuclear antigen (ENA)
EIA
Anti-smooth muscle antigen (ASMA)
IFI
Anti-neutrophil cytoplasmic (ANCA)
IFI
Anti-myeloperoxidase (MPO)
EIA
Anti-proteinase 3 (PR3)
EIA
Anti-tissue transglutaminase (t-TG)
IgA-EIA
Anti-endomysial (EMA)
IgA-IFI
Anti-thyroglobulin
CLIA
Anti-thyroid peroxidase (TPO)
CLIA
Anti-cardiolipin
ELISA
Anti-liver-kidney microsomial (LKM)
IFI
Anti-Mi-2, Ku, PM-Scl100, PM-Scl75, Jo-1,
SRP, PL-7, PL-12, EJ, OJ, Ro-52, Scl-70,
CENP A, CENP B, RP11, RP155, Fibrillarin,
NOR90, Th/To, PDGFR
EUROLINE
Anti-Ach receptor
RIA
Negative
Negative
Anti-ganglionic Ach receptor (α3-α7)
Neuronal surface-exposed antigens
Neuronal intracellular antigens
Other neuronal antigens
Anti-NMDA receptor
transfection
Anti-AMPA 1,2 receptor
transfection
Anti-GABA B receptor
transfection
Anti-Myelin associated glycoprotein (MAG)
IgM-ELISA
Anti-GD1a, b; GM1, 2, 3; GQ1b, GT1b
transfection
Anti-VGKC (LGI 1, CASPR 2)
transfection
Anti-amphiphysin
Rec. protein
Anti-Hu, Yo, Ri, CV2, Ma2/Ta, GAD
Rec. protein
Negative
Anti-basal ganglia
H. brain IB
Positive
Anti-α enolase
Rec. protein IB
Negative
Anti-γ enolase
Rec. protein IB
Positive
Anti-pyruvate kinase
Rec. protein IB
Negative
19
Negative
20