“Chronic Lyme Disease Validated” by Discover Magazine Selected in Top 100 News Stories of 2011.

It is time to take their collective heads out of the sand…yes, that’s right; the   “play it safe” doctors, the “lets not make any waves” doctors, the “militaristic” and “opinionated” doctors…As I have been predicting, the facts have caught up with those in power, and it will be very interesting to see how they try to ignore the new evidence.

You can bet Dr. S and others will be trying to discredit this study:

Voted as one of the: Top 100 Stories in Discover Magazine in 2011, “Chronic Lyme Patients Validated” Patients with chronic fatigue syndrome and post-treatment Lyme disease syndrome (in which symptoms persist after antibiotic treatment) have spent decades fending off charges that their debilitating exhaustion and cognitive problems were simply imagined. But a study released last February provides tangible evidence that their conditions are real and distinct entities.

Immunologist Steven Schutzer of the University of Medicine and Dentistry of New Jersey examined samples of cerebrospinal fluid, the clear liquid aE¨that surrounds the brain and spinal cord, from patients with each syndrome. In identifying the contents of that fluid, he documented different sets of proteins for each group of patients, potential biomarkers that distinguish between the two aE¨conditions and healthy controls.

Schutzer revealed the marker proteins by removing common, unrelated proteins like albumin and immunoglobulin from the spinal fluid before his analysis. aEoeThat lets the smaller proteinsaE”the potential biomarkersaE”not get obscured,aE he says.

aEoeAt least now we know weaE™re not just speculating about the differences between chronic fatigue syndrome and post-treatment Lyme.aE

The full article follows from his publication in PubMed:

Distinct Cerebrospinal Fluid Proteomes Differentiate Post-Treatment Lyme Disease  from Chronic Fatigue Syndrome

Steven E. Schutzer1*., Thomas E. Angel4., Tao  Liu4., Athena A. Schepmoes4, Therese R. Clauss4, Joshua N. Adkins4, David G. Camp  II4, Bart  K. Holland3, Jonas Bergquist5, Patricia K. Coyle6, Richard D. Smith4, Brian  A. Fallon7, Benjamin H. Natelson2,8

1 Department of Medicine, University of Medicine and Dentistry of New Jersey-New Jersey Medical School, Newark, New Jersey, United States of America, 2 Department of Neurology, University of Medicine and Dentistry of New Jersey-New Jersey Medical School, Newark, New Jersey, United States of America, 3 Division of Biostatistics and Epidemiology, University of Medicine and  Dentistry of New Jersey-New Jersey Medical School, Newark, New Jersey, United States  of America, 4 Biological Sciences Division, Pacific Northwest National Laboratory, Richland, Washington, United States of America, 5 Department of Physical and Analytical Chemistry, Uppsala University, Uppsala, Sweden, 6 Department of Neurology, State University of New York-Stony Brook, Stony Brook, New York, United States of America, 7 Department of Psychiatry, Columbia University Medical Center, New York, New York, United States of America, 8 Department of Pain Medicine and Palliative Care and Beth Israel Medical Center, Albert Einstein School of Medicine, Bronx, New York, United States of America.

Introduction

Prime objectives in studying neurologic and psychiatric disorders are  to  develop discriminating markers and generate data that can provide insight into disease pathogenesis. This can lead to novel treatment strategies. Chronic  Fatigue Syndrome (CFS) and Lyme disease, particularly Neurologic Post Treatment Lyme disease syndrome (nPTLS), represent two conditions that share common symptoms of fatigue and cognitive dysfunction [1aE”7].

Despite extensive research CFS and nPTLS remain medically unexplained. There are no biological markers to distinguish these syndromes, creating diagnostic dilemmas and impeding research into understanding each individual syndrome. Cerebrospinal fluid (CSF) is an ideal body fluid to examine for signature protein profiles informative for diagnosis or etiology of central nervous system (CNS)-related symptoms and dysfunction.

Not only is the CSF an accessible liquid extension of the brain, but recent data suggests CSF may provide more relevant data than brain parenchyma itself in certain neurologic diseases [8]. Specific abnormalities found in CSF relating to CFS and  nPTLS would suggest CNS involvement, and could facilitate their mechanistic understanding.

Liquid chromatography coupled to mass spectrometry (LC- MS) is becoming the method of choice for examining complex biological specimens, that  contain  hundreds  to  thousands   of proteins [9], such as CSF [10]. This is particularly the case in the initial discovery phase. This  discovery  phase may be   viewed as casting a wide net to maximize identification of as many proteins as possible in a sample. This initial list of identified proteins has value by itself for qualitative or   semi-quantitative comparisons between diseases. Recent studies demonstrated the reliability and reproducibility of LC-MS results, with different mass spectrometers across different  laboratories, when   performed  by experienced individuals [9,11].

In a discovery phase investigation, the MS technique is unbiased and does not require prior knowledge of what  proteins   may  be  in  a  sample. This  is in  contrast  to subsequent validation studies where targeted approaches are used and which do require knowledge of target proteins. In searching for a disease biomarker, the discovery phase should provide a list of proteins and serve as a precursor phase for targeted approaches. These subsequent targeted approaches, whether they use other  MS  techniques   or  are  immuno-based,  are  designed to validate the   use of the   biomarker protein(s) to distinguish one disease from another.

In  practice tailored strategies are often needed   to achieve a balance  between  ideal   and  real  world constraints   aE”  especially where sample volumes and numbers are limited such as with CSF. In an ideal situation it is desirable to have numerous samples from individuals with a particular disease. It is further desirable to have sufficient total protein content in each sample so that a variety of protein separation and fractionation methods can be used prior to MS analysis. This will minimize abundant proteins from masking the detection of less abundant ones, and will permit full qualitative and quantitative analyses. Limited sample numbers and quantities do not preclude employment of tailored strategies to get meaningful results. It should be remembered that in the example of a biomarker search, the protein(s) will be confirmed or dismissed in future targeted validation studies, but failure to identify them in the broad discovery list would preclude them from examination for validation.

Until recently, technical hurdles impeded the use of CSF to distinguish conditions such as CFS and nPTLS. Advances in sample preparation, separations and MS platform capabilities enabled us to recently establish a  comprehensive reference normal CSF proteome [10]. This provides the basis for comparative proteome analyses with other diseases, which should provide greater insight into their underlying pathogenesis.

To address the possibility that CFS and nPTLS could be distinguished from  one another  and healthy subjects, we searched for distinguishing protein marker profiles by applying our advanced proteomics strategy [10] to characterize the CSF proteomes from well described CFS and nPTLS patients (detailed in Methods). We performed comparative whole CSF proteome analyses between CFS, nPTLS, and healthy normal controls, and complemented these findings with label-free quantitative analysis of individual subject samples. In addition, we performed a preliminary pathway analysis [12] using these data, to examine the  feasibility  of this   type  of tool  for   future  investigations to probe  for  clues to  the  pathogenetic  mechanisms behind  these diseases.

Materials and Methods

Ethics Statement

Approval for the conduct of this study was obtained from the Institutional Review Board of New Jersey Medical School and the Institutional Review Board of Pacific Northwest National Laboratory (Exempt status and consent not required, using previously banked de- identified samples in accordance with federal regulations).

Overview and Rationale

We performed analysis of pooled CSF samples allowing for a broad   and deep view as well as qualitative comparison of each disease-related and control CSF proteome. To determine if these two  syndromes   could  be  quantitatively  differentiated we  per- formed a label-free quantitative analysis of protein abundances for individual subject CSF samples. Pooling samples provided sufficient protein  mass for effective downstream proteomics analysis   following immunoaffinity  depletion  of  the  14  most abundant  proteins present   (representing approximately   95%  of the total protein mass in CSF), reducing the dynamic range of protein  concentrations  present  in  CSF,  where   proteins   with highest concentrations mask proteins at lower concentrations from detection. Coupling immunoaffinity depletion with strong cation exchange (SCX) fractionation further reduces sample complexity, and allowed for the in-depth analysis of the CSF proteomes. These comprehensive CSF proteomics datasets were then used to create an accurate mass and time (AMT) tag database for subsequent label-free quantitative analysis of individual subject CSF samples. Due  to the  limit in sample volume, the  CSF   samples used in individual LC-MS analyses were not immunoaffinity depleted and fractionated, and therefore had much lower proteome coverage compared  to  the  pooled   samples. Nevertheless, the  label-free quantitative analysis of single subject samples provided a means for statistical evaluation  of the  quantified   protein  abundances  for many subjects suffering from CFS and nPTLS as well as normal healthy volunteers. Together these analyses represent the discovery phase of our studies on CFS and nPTLS, generating targets for follow up  verification and  validation in the  later   stages of the biomarker discovery workflow [13].

 Cerebrospinal Fluid (CSF) specimens

CFS  Subjects. Both pooled and individuals CSF  samples were analyzed. Equal aliquots from individual CSF samples were pooled to provide sufficient volume for extensive fractionation and two-dimensional LC  coupled to   tandem  MS  (2D-LC-MS/MS) analysis with immunoaffinity depletion from 30 women and 13 men (n = 43) who fulfilled the 1994 case definition for CFS [1]. All subjects were 18aE”54 years old (median  = 43) and   underwent a careful history and physical examination by an expert experienced in evaluating patients  with medically unexplained  fatigue and pain. Patients had blood tests to rule out common causes of severe fatigue such as anemia,   liver disease, hypothyroidism, systemic lupus erythematosus, and   Lyme disease [14]. All subjects then underwent a psychiatric diagnostic interview designed to identify major psychiatric diagnoses for exclusion in this study. Eleven of the patients were not taking medicines. Subjects then underwent lumbar puncture. CSF was sent to the laboratory for white blood cell (wbc) count and total protein [10]. A majority of CFS patients had normal CSF protein and cell counts (protein less than 45 mg/dl and wbc less than or equal to 5/mm3). Ten of the patients had increased protein values ranging from 46aE”93 mg/dl (with   a median of 59 mg/dl) and 3 patients had minimally elevated wbc counts of 6, 7, and 9 respectively. Individual CSF samples from 14 of the 43 CFS subjects (aged 33aE”48 years with a median age of 43

years, 7   female and  7   male) were also used in   direct  LC-MS analysis (i.e., no MS/MS  was performed) without immunoaffinity depletion. Twelve of the  14 patients   had  normal  CSF  protein levels and all had normal cell counts. All subjects provided written informed consent approved by the Institutional Review Board.

nPTLS Subjects.    Both pooled and individuals CSF samples were analyzed. Equal  aliquots from individual CSF   samples were pooled to provide sufficient volume for  extensive frac- tionation  and  2D-LC-MS/MS    analysis with   immunoaffinity depletion from  15 females and 10 males (n = 25) with nPTLS. All were documented to have had   prior Lyme disease which met CDC  surveillance case definition criteria [15], persistent neurologic features, including cognitive impairment and fatigue, despite appropriate antibiotic treatment [16,17]. Sub- jects were 17aE”64 years old (median = 48). All were seropositive for   antibodies to  B. burgdorferi  (the etiologic agent  of Lyme disease). Patients, enrolled in an NIH  funded study, met the following criteria [17]: (1)  current  positive IgG Western blot using CDC  surveillance criteria assessed using a single reference laboratory (University  Hospital of Stony Brook); (2) treatment for Lyme disease with at least 3 weeks of intravenous ceftriaxone or cefotaxime that was completed at least 4 months before study entry; and (3) objective evidence of memory impairment as documented by the Wechsler Memory Scale-III compared to age-, sex-and education-adjusted population norms.  nPTLS subjects were  excluded  if  history   or  testing revealed a medical condition that could cause cognitive impairment  or confound neuropsychological assessment (e.g., neurological   disease, autoimmune disease, unstable  thyroid disease, learning disability, substance abuse, B12 deficiency). Patients with cephalosporin allergy or a history of significant psychiatric disorder prior to onset of Lyme disease were also excluded.

All patients had   a comprehensive battery of neur-cocognitive testing and a full-physical exam with detailed rheumatologic  and  neurologic   assessments. nPTLS patients then had a lumbar puncture   and CSF was evaluated for cell count, total protein, glucose, total gammaglobulin, oligoclonal bands and evidence of B. burgdorferi (ELISA, Bb DNA by PCR, and culture using BSKII medium). None had evidence of another active tick-borne disease. A majority of nPTLS patients included in the pooled sample had  normal  CSF protein and cell counts (protein less than  45 mg/dl  and   wbc less than  or equal  to  5/mm3),  except  for   3  patients  who  had  elevated protein  values of  58,  69,  and  71 mg/dl     respectively and  1 patient with elevated wbc count of 6. Individual CSF samples from a group of 14 of the 25 nPTLS subjects (aged 25aE”58 years with a median age of 48 years, 6 female and 8 male) were also used in direct LC-MS analysis without immunoaffinity depletion. Two of the 14 patients had increased CSF protein levels of 69 and 71 mg/dl  and 1 had a slightly elevated wbc of 6. All subjects provided written informed consent approved by the Institutional Review Board.

Normal  Controls. We used the 2D-LC-MS/MS data obtained previously from pooled CSF of 11 healthy  control subjects [10]. Briefly, there were 8 women and 3 men, aged 24aE”55 years with a median age of 28 years. Individual CSF samples from another set of 10 healthy volunteers, age 37aE”44 years (median = 40) and 5 women and 5 men, were analyzed by LC-MS analysis without immunoaffinity depletion.

Immunoaffinity depletion of 14 high abundance CSF proteins. We had previously shown that this technique could increase our protein identification yield by 70% [10]. Pooled CSF samples from CFS or  nPTLS patients (total   volume  of  18 mL  each),  were fractionated using a 12.7679.0 mm SepproH IgY14 LC10 affinity LC column (Sigma, St Louis, MO) as previously described [18]. Pooling was done to compensate for lack of sufficient volume (and consequent protein content) available for immunoaffinity deple- tion of individual patient samples. Both the flow-through (lower abundance proteins) and bound fractions from both pooled CSF samples were collected and processed identically until LC-MS/MS analysis. These analyses resulted in an in-depth characterization of the CSF proteome and the combined results of abundant protein and   less abundant  protein fractions allowed the creation of an AMT tag database [19] for high-throughput analysis of a larger number of individual subject samples using LC-MS.

Protein digestion

CSF proteins  (from the immunoaffinity depletion   processed pooled samples and the individual samples without immunoaffinity depletion processing) were digested with trypsin and cleaned up   with SPE C18  columns as previously described [10]. Final peptide concentration was determined by  BCA assay (Pierce, Rockford,  IL).   All tryptic digests were  snap  frozen  in  liquid nitrogen  and  stored at 280uC until further processing and analysis.

Strong cation exchange  (SCX) fractionation

A total of 300 mg of tryptic peptides from both the IgY14 bound and flow-through fractions from the pooled CFS and nPTLS CSF samples were fractionated by SCX chromatography as described [20]. Thirty  SCX fractions were collected for each sample and 20% of each fraction was injected for reversed-phase LC-MS/MS analysis.

 Reversed-phase  capillary LC-MS/MS for CSF pooled fraction analysis SCX  fractions  of  the  IgY14  bound  fraction  samples were analyzed on an LTQ  (ThermoFisher, San Jose, CA) linear ion trap, and SCX fractions of the IgY14 flow-through fraction samples were   analyzed  on  an  LTQ-Orbitrap  Velos (Thermo- Fisher) instrument,  operated  in data-dependent  mode   with the same LC conditions as previously described [10].

Reversed-phase  capillary LC-MS for label-free quantification  of unfractionated CSF samples   (individual  patient     samples  with  insufficient volume (protein content) for immunoaffinity depletion and SCX fraction- ation), the LTQ-Orbitrap Velos mass spectrometer was operated in the data-dependent mode with full scan MS spectra (m/z 400aE” 2000) acquired  in  the  LTQ-Orbitrap Velos with resolution of 60,000 at m/z 400 (accumulation target: 1,000,000). MS/MS data acquired here were not used for the quantitative analysis.

Data analysis

The  LTQ  raw data  from the   pooled samples was extracted using Extract_MSn (version 3.0; ThermoFisher) and analyzed with the SEQUEST  algorithm (V27 revision 12; ThermoFisher) searching the MS/MS  data against the human IPI database (Version 3.40). Mass tolerances of 3 Daltons for precursor ions and 1 Dalton for fragment ions without an enzyme defined, as well as static carboxyamidomethylation of cysteine and dynamic oxidation of methionine were used for the database search. The LTQ- Orbitrap  Velos MS/MS  data   were first processed by in-house software DeconMSn [21] accurately determining the monoisotopic mass and charge state of parent ions, followed by SEQUEST search against the IPI database in the same fashion as described above, with the exception that   a 0.1-Dalton mass tolerance for precursor ions and 1-Dalton mass tolerance for fragment ions were used. Data   filtering criteria based on the cross correlation score (Xcorr) and  delta  correlation  (DCn)  values along   with  tryptic cleavage and charge states were developed using the decoy database approach and applied for filtering the raw data to limit false positive identifications to ,1% at the peptide level [22aE”24]. For the LTQ-Orbitrap Velos data, the distribution of mass deviation (from the  theoretical masses) was first determined  as having a standard deviation (s) of 2.05 part per million (ppm), and a mass error of smaller than   3s was used in combination with Xcorr and DCn to determine the filtering criteria that resulted in 1% false positive peptide identifications.

The AMT tag strategy [19] was used for label-free quantification of MS features observed in the LTQ-Orbitrap Velos analysis of the individual CSF samples from normal, CFS and nPTLS conditions. The filtered MS/MS  peptide identifications obtained from the 2D- LC-MS/MS analyses of all pooled CSF samples were included in an AMT tag database with their theoretical mass and normalized elution time (NET;   from   0   to  1)   recorded.  LC-MS  datasets were then analyzed by in-house software VIPER [25] that detects features in massaE”NET space and assigned them to peptides in the AMT tag database [26]. The   data was further filtered by requiring that all peptides must be detected in at least 30% of the datasets in each of the three conditions. The false discovery rate of the AMT tag analysis was estimated using an 11-Da shift strategy as previously described [27]. A false positive rate of ,4% was estimated for each of the LC-MS data sets. The  resulting lists of peptides from 2D-LC-MS/MS  or direct LC-MS analysis were further processed by ProteinProphet software [28] to remove redundancy in protein identification.

Data normalization and quantification of the changes in protein abundance  between the normal,  CFS and  nPTLS CSF samples were performed and visualized using in-house software DAnTE [29]. Briefly, peptide intensities from the LC-MS analyses of the individual samples (volume limited) were log2 transformed and normalized  using a  mean  central  tendency   procedure.  Peptide abundances from the individual samples were then aE˜aE˜rolled upaE™aE™ to the protein level employing the R-rollup method (based on trends at  peptide  level) implemented in DAnTE. ANOVA, principal component  analysis (PCA) and  clustering  analyses   were  also performed using DAnTE.

Pathway Analysis of the   data  was performed with Ingenuity Pathways Analysis (Ingenuity Systems, www.ingenuity.com).Canonical  pathway  analysis identified the  pathways   from  the Ingenuity Pathways Analysis library of canonical pathways that were most significant to the CFS and nPTLS proteins identified. The significance of the associations were assessed with the FisheraE™s exact test.

Results

We  first performed  pooled   sample analysis, then  individual sample analysis, and   then   pathway analysis using the observed proteins. These analyses represent a discovery phase of our studies on CFS and nPTLS, generating targets which can be followed up in future verification and validation stages studies [13].

Proteomic analysis of pooled  CSF samples

In  the   pooled analysis, we examined individual sets of CSF samples from CFS patients (n = 43) and nPTLS patients (n = 25), respectively. We used the proteomic strategy described in Methods to  assure  that  the  maximum  number  of  proteins  would  be analyzed and the more abundant proteins did not obscure the less abundant  ones having biomarker potential. The  bound  fraction of  abundant     proteins  from  the  immunoaffinity depleted  flow through  fraction  was analyzed   separately and  included   in  the subsequent analysis. Combining immunoaffinity-based partition- ing, SCX fractionation and LC-MS/MS,  we identified approximately 30,000 peptides for each pooled sample corresponding to 2,783 non-redundant   proteins in CFS patient samples and 2,768 proteins in nPTLS patient samples, compared to the 2,630 proteins present in the CSF of healthy normal control subjects. These can be  graphically seen   in  Figure   1  which  shows the  number  of proteins identified solely in each group, and shared or not shared between the groups (see Table S1). Figure 1 also shows that the nPTLS and CFS groups shared significantly more proteins (n = 305) than  each disease group shared with healthy controls (naE™s = 135 and   166, respectively). (Note that,  as with any assay, when we indicate that a protein was aE˜aE˜not foundaE™aE™ or aE˜aE˜not identifiedaE™aE™ that is defined as within the limits of detection).

Compared  to 10 normal  healthy volunteers (samples chosen at random) to provide insights on the variation among individuals within   and  between  different groups.  Limited  volumes of the individual samples reduced the sample preparation  options (i.e., immunoaffinity depletion  and  SCX  fractionation),   and  hence resulted in less depth of proteome coverage than possible with the pooled samples, where approximately 20 ml were available for depletion  and  fractionation.  Nevertheless, we   identified 4,522 peptides across all individual samples, representative of 474 non- redundant  proteins   identified and  quantified   in  the  individual sample analysis (Table S2).

Unsupervised hierarchical clustering and PCA were employed to determine if  the observed quantitative differences in protein abun- dances were sufficient to distinguish these two patient groups (this was de facto blinded aE” as samples were run in a random order and uncoded as  to  disease group  afterwards). The  proteins  considered   in  the unsupervised hierarchical clustering analysis were quantified in individual samples and found to be significantly different in abundance by analysis of variance (ANOVA p # 0.01, Table S3); while PCA analysis considered all proteins quantified in each individual sample. The CSF proteome of the two disease states were markedly different from each other (Fig. 2A and   B).  Individual patients also showed consistent patterns of protein abundances discriminating CFS from nPTLS (Fig. 2A). These results demonstrated that it is unlikely that any single subjectaE™s CSF  sample   in  the  pooled  analysis contributed disproportionately to the differential proteome distributions observed between the disease groups. Moreover, the individual analyses also highlighted the potential for diagnostic marker confirmation upon extension to larger sample sets in validation studies.

Illustrative pathway  analyses of protein  results from CSF samples

We utilized pathway analysis as an exploratory tool to assess the value of our data, beyond distinguishing the two syndromes from each other, to see if the data was amenable to analysis that would help generate hypotheses of pathogenesis. We chose representative pathways to analyze for illustration based in part on their quantitative  ranking  (Table  S4) and  in  part  by   the  potential relevance of the pathway involved. Even this limited investigation demonstrated that there is a wealth of proteome information that can be leveraged for hypotheses generation.

Example    of    proteins     in      common  and    elevated  in abundance   in    the    two    disease  conditions,  compared  to normal, but    at different levels.     An illustration, where the same proteins are elevated in abundance in both conditions, but at different magnitudes, is provided by inspection of proteins in the complement system. This is of interest because both syndromes may be triggered by infections (nPTLS in all cases by B. burgdorferi; many CFS cases by one or more microbes yet to be identified). We found that the complement cascade  related proteins were identified and  significantly enriched in both CFS and  nPTLS pooled  CSF  proteomes  by  the  Fisher  Exact  test  (p = 0.005) implemented  in  Ingenuity   Pathways Analysis (Figure S1A). In individual patient samples analyzed, we identified and quantified 4 components  (C1S, C4B,  C1QB,  C1QC)  which   are  seen with activation of the complement cascade and which were differentially increased in abundance  consistently across the nPTLS patients  compared  to  CFS (Figure S1B   and  C).  This represents the type of data that can be useful in the formulation of pathogenetic hypotheses because the role of complement in these disorders is under-explored.

Example of proteins solely identified in one condition. Analysis of the  highly fractionated   pooled patient samples led to the identification of proteins solely identified in each of the disease states. To investigate if these disease specific proteins have common annotated functional properties, we performed pathway analysis (Tables S5 and S6). As an example, the CDK5 signaling  pathway,    was  found    to    be    significantly  enriched (p = 0.00009)  for  proteins  identified  only  in the pooled  CFS proteome. This signaling pathway has been linked to ParkinsonaE™s [30] and AlzheimeraE™s diseases [31].

Example of proteins in  common and decreased in abundance  in     the    two    disease  conditions,  compared  to normal, but  at different levels.    In certain cases, proteins were found  to  be  decreased   in  both  CFS and nPTLS compared  to healthy normal controls. However, quantitative distinguishing differences could still be found between the   two conditions. A specific example  relates  to  networks   relevant  to  neurological function such as axonal guidance (Figures S2A and B), where the proteins in CFS were further decreased relative to nPTLS. These findings highlight quantifiable differences between CFS and nPTLS that may be found, with respect to certain proteins such as those that   are known to effect the dynamic changes in CNS cellular architecture, such as axon, neurite, and dendritic spine growth and organization.

Discussion

Our  results support  the  concept   that  CFS and  nPTLS are distinguishable disorders with distinct CSF proteomes, where one can be separated from the other. The results also demonstrate that each condition has a multitude of candidate diagnostic biomarkers for future validation and optimization studies. The   discovery of many of the same proteins in each proteome is important because it allows comparative pathway analysis, so that useful hypotheses of pathogenesis can be formulated and tested.

Our  results represent the most comprehensive analysis of the whole CSF proteome to date for both CFS and nPTLS. These two disorders have   similar symptoms that  have   created  diagnostic dilemmas. It has been speculated that one (nPTLS) is a subset of the other, but our results do not support that notion. Our findings alone do not describe why CFS or nPTLS occur, but are provided to illustrate that CSF proteome analysis may provide important and meaningful insights into the biological processes modulated as a function of disease and   facilitate the identification of protein candidates for further investigation. Analytical strategies need to be developed for application to those proteins and their pathways that may not have been described yet. Nevertheless, in toto, these results are encouraging because there is an abundance of data now that   can be analyzed with existing tools and future methods to develop hypotheses on pathogenesis [9,32].

We regard the proteins that were identified only in one group or differentially abundant  between groups, as possible or candidate biomarkers that can be subjected to further analysis in validation and verification studies. The  clinical significance of the proteins identified in each pooled sample is difficult to determine in the current   discovery phase. As with most technological methods, we expect multiple replicate analyses of the highly fractionated samples would result in a reduction of the number of seemingly unique proteins identified for each disease group [33].

An important strategy that can be used post-discovery towards validation, is the use of targeted approaches that are either MS- based, immuno-based, or a combination of these approaches [12,34]. One approach, selected reaction monitoring (SRM) MS, allows for much higher sensitivity and specificity, more accurate quantification, and much higher throughput   to be achieved for simultaneously measuring many  biomarker  candidates  in   large clinical cohorts [35aE”37]. This approach also compensates for any theoretical over-representation of proteins in pooled samples by a single or small number of individuals.

This is a strategy that we plan to use not only for these diseases, but in the investigation of other diseases with neuropsychiatric features. SRM-MS analysis will permit  us to  directly use small-sized samples, such as the individual CSF samples, enable verification of marker candidates that currently do not have available antibodies (hence not amenable  to conventional analyses such as ELISA or   Western blots), and provide robust statistical analyses on individual candidate markers or combinations of them to determine which would make the best biomarker(s) for a particular disease condition. Immunobased assays such as ELISA or Western blots may also be used for targeted approaches, but will likely have more utility during a clinical validation phase where much larger sample cohorts are used. Some may choose to apply these methods for additional  orthogonal  confirmation  of   a  result.  However,  its greater value may lie in its widespread use as a common diagnostic platform. Regardless of the method chosen, identification of diagnostic CSF biomarkers may be the necessary prelude to a search for the same markers in the highly complex blood, because it permits targeted searches for markers that might otherwise be obscured or have uncertain relevance.

With respect to biomarkers, we believe our proteomic strategy [10], that did not require prior knowledge of which proteins might be  present  in  the  CSF,  will accelerate   the  transition  from  a discovery phase  of candidate  biomarkers,   as   described  in  this study, to full validation for clinical application. We and   others have cited important elements that should be considered when an assay or  biomarker  is being developed for preliminary   or  full validation [38aE”40].

Distinguishing CFS and nPTLS will have etiologic implications which could lead to novel diagnostics and therapeutic interventions. On  a broader   level the strategy we employed may prove useful   in  providing  investigative foundations  in  other  poorly understood neurological conditions.

Author Contributions

Conceived and designed the experiments: SES TEA TL RDS. Performed the experiments: SES TEA TL AAS TRC.   Analyzed the data: SES TEA TL   JNA RDS   PKC  BAF BHN DGC  BKH JB. Contributed  reagents/ materials/analysis tools: SES TEA TL JNA RDS PKC BAF BHN DGC. Wrote the paper: SES TEA TL JNA RDS PKC BAF BHN DGC.

 

 

 

References

 

1.  Fukuda K, Straus SE, Hickie I, Sharpe MC,  Dobbins JG, et al. (1994) The chronic fatigue syndrome: a comprehensive approach to its definition and study. International  Chronic Fatigue Syndrome Study Group. Ann Intern Med 121: 953aE”959.

2.  Natelson  BH,  Weaver  SA,   Tseng  CL,  Ottenweller  JE  (2005) Spinal  fluid abnormalities in patients with chronic fatigue syndrome. Clin Diagn Lab Immunol 12: 52aE”55.

3.  Coyle PK (1992) Neurologic Lyme disease. Semin Neurol 12: 200aE”208.

4.  Pachner AR, Steere AC (1985) The triad of neurologic manifestations of Lyme disease: meningitis, cranial neuritis, and radiculoneuritis. Neurology 35: 47aE”53.

5.  Logigian EL,   Kaplan  RF,  Steere   AC   (1999) Successful treatment  of  Lyme encephalopathy with intravenous ceftriaxone. J Infect Dis 180: 377aE”383.

6.  Logigian EL, Kaplan RF, Steere AC (1990) Chronic neurologic manifestations of Lyme disease. N Engl J Med 323: 1438aE”1444.

7.  Gaudino  EA, Coyle PK, Krupp  LB (1997) Post-Lyme syndrome and chronic fatigue syndrome. Neuropsychiatric similarities and differences. Arch Neurol 54: 1372aE”1376.

8.  Ransohoff RM  (2009) Immunology: Barrier to electrical storms. Nature  457: 155aE”156.

9.  Aebersold R (2009) A stress test for mass spectrometry-based proteomics. Nat Methods 6: 411aE”412.

10.  Schutzer SE, Liu T,   Natelson BH, Angel TE,  Schepmoes AA, et al. (2010) Establishing the proteome of normal human cerebrospinal fluid. PLoS ONE 5: e10980.

11.  Addona TA, Abbatiello SE, Schilling B, Skates SJ, Mani DR, et al. (2009) Multi- site  assessment of     the  precision  and  reproducibility  of  multiple  reaction monitoring-based  measurements  of proteins  in  plasma.  Nat  Biotechnol 27: 633aE”641.

12.  Schmidt  A,  Claassen  M,  Aebersold R  (2009) Directed  mass spectrometry: towards hypothesis-driven proteomics. Curr Opin Chem Biol 13: 510aE”517.

13.  Surinova S, Schiess R, Huttenhain R, Cerciello F, Wollscheid B, et al. (2011) On the development of plasma protein biomarkers. J Proteome Res 10: 5aE”16.

14.  Schutzer  SE,  Natelson  BH  (1999)  Absence of   Borrelia   burgdorferi-specific immune complexes in chronic fatigue syndrome. Neurology 53: 1340aE”1341.

15.  CDC  (1997) Case  definitions for   infectious conditions under  public   health surveillance. Centers for Disease Control and Prevention. MMWR Morb Mortal Wkly Rep 46: 1aE”55.

16.  Wormser GP, Dattwyler RJ, Shapiro ED, Halperin JJ, Steere AC, et al. (2006) The  clinical assessment, treatment,  and  prevention   of lyme disease, human granulocytic anaplasmosis, and  babesiosis: clinical practice guidelines by the Infectious Diseases Society of America. Clin Infect Dis 43: 1089aE”1134.

17.  Fallon BA, Keilp JG, Corbera  KM,  Petkova E, Britton CB, et al. (2008) A randomized, placebo-controlled trial of repeated IV antibiotic therapy for Lyme encephalopathy. Neurology 70: 992aE”1003.

18.  Liu   T,  Qian  WJ, Mottaz  HM,  Gritsenko MA,  Norbeck  AD,   et  al.  (2006) Evaluation of multiprotein immunoaffinity subtraction for plasma proteomics and candidate biomarker discovery using mass spectrometry. Mol Cell Proteomics 5: 2167aE”2174.

19.  Smith RD, Anderson GA, Lipton MS, Pasa-Tolic L, Shen Y, et al. (2002) An accurate  mass tag  strategy for   quantitative  and  high-throughput  proteome measurements. Proteomics 2: 513aE”523.

20.  Liu T,  Qian  WJ, Gritsenko MA, Xiao W, Moldawer LL, et al. (2006) High dynamic range characterization of the trauma  patient plasma proteome. Mol Cell Proteomics 5: 1899aE”1913.

21.  Mayampurath AM, Jaitly N, Purvine SO, Monroe ME, Auberry KJ, et al. (2008) DeconMSn: a software tool for accurate parent ion monoisotopic mass determination for tandem mass spectra. Bioinformatics 24: 1021aE”1023.

22.  Elias JE, Gygi SP (2007) Target-decoy search strategy for increased confidence in  large-scale protein  identifications by mass spectrometry. Nat  Methods  4: 207aE”214.

23.  Peng J, Elias JE, Thoreen  CC,   Licklider LJ, Gygi SP (2003) Evaluation of multidimensional chromatography  coupled  with   tandem  mass spectrometry (LC/LC-MS/MS)    for  large-scale  protein  analysis:  the  yeast  proteome. J Proteome Res 2: 43aE”50.

24.  Qian  WJ,   Liu  T,  Monroe  ME,  Strittmatter  EF,  Jacobs JM,  et  al.  (2005) Probability-based evaluation of peptide and protein identifications from tandem mass spectrometry and SEQUEST  analysis: the human proteome. J Proteome Res 4: 53aE”62.

25.  Monroe ME, Tolic N, Jaitly N, Shaw JL, Adkins JN, et al. (2007) VIPER: an advanced software package to support high-throughput LC-MS peptide identification. Bioinformatics 23: 2021aE”2023.

26.  Zimmer JS, Monroe ME, Qian WJ, Smith RD (2006) Advances in proteomics data analysis and display using an accurate mass and time tag approach. Mass Spectrom Rev 25: 450aE”482.

27.  Petyuk VA, Qian  WJ, Chin MH,   Wang H, Livesay EA, et al. (2007) Spatial mapping of protein abundances in the mouse brain by voxelation integrated with high-throughput liquid chromatography-mass spectrometry. Genome Res 17: 328aE”336.

28.  Nesvizhskii AI, Keller A, Kolker E, Aebersold R (2003) A statistical model for identifying proteins by tandem mass spectrometry. Anal Chem 75: 4646aE”4658.

29.  Polpitiya AD, Qian WJ, Jaitly N, Petyuk VA, Adkins JN, et al. (2008) DAnTE: a statistical tool  for  quantitative  analysis of  -omics  data.  Bioinformatics 24: 1556aE”1558.

30.  Smith PD, Crocker SJ, Jackson-Lewis V, Jordan-Sciutto KL, Hayley S, et al. (2003) Cyclin-dependent kinase 5 is a mediator of dopaminergic neuron loss in a mouse  model  of  ParkinsonaE™s  disease.  Proc  Natl  Acad  Sci  U  S  A  100: 13650aE”13655.

31.  Lew J, Huang  QQ,  Qi Z, Winkfein RJ, Aebersold R, et al. (1994) A brain- specific activator of cyclin-dependent kinase 5. Nature 371: 423aE”426.

32.  Domon  B, Aebersold R  (2010) Options  and  considerations when selecting a quantitative proteomics strategy. Nat Biotechnol 28: 710aE”721.

33.  Mitchell P (2010) Proteomics retrenches. Nat Biotechnol 28: 665aE”670.

34.  Pan  S,  Aebersold R,  Chen  R,  Rush  J,   Goodlett  DR,  et  al.  (2009) Mass spectrometry based targeted protein quantification: methods and applications. J Proteome Res 8: 787aE”797.

35.  Rifai  N,  Gillette  MA,  Carr     SA  (2006) Protein  biomarker  discovery and validation: the long and uncertain path to clinical utility. Nat Biotechnol 24: 971aE”983.

36.  Schiess R, Wollscheid B, Aebersold R (2009)  Targeted proteomic strategy for clinical biomarker discovery. Mol Oncol 3: 33aE”44.

37.  Hossain M, Kaleta DT, Robinson EW, Liu T, Zhao R, et al. (2010) Enhanced sensitivity for selected reaction  monitoring-mass spectrometry-based targeted proteomics using a dual-stage electrodynamic ion funnel interface. Mol Cell Proteomics, Epub ahead of print.

38.  Budowle B, Schutzer SE, Morse SA, Martinez KF, Chakraborty R, et al. (2008) Criteria  for  validation  of   methods  in  microbial  forensics. Appl   Environ Microbiol 74: 5599aE”5607.

39.  Mischak H, Allmaier G, Apweiler R, Attwood T,   Baumann M, et al. (2010) Recommendations  for   biomarker  identification and  qualification in  clinical proteomics. Sci Transl Med 2: 46ps42.

40.  Schutzer SE, Keim P, Czerwinski K, Budowle B (2009) Use of forensic methods under exigent circumstances without full validation. Sci Transl Med 1: 8cm7.

 

 

 

 

 

(Visited 27 times, 1 visits today)
Subscribe to Jenna's Lyme Blog
Yes, I want to subscribe. I understand I will only receive one email each month when there are new posts.
This entry was posted in Chronic Lyme Disease, Diagnosis of Lyme Disease, Lyme Disease News, Neurological Lyme disease, Research and Development and tagged , , . Bookmark the permalink.

2 Responses to “Chronic Lyme Disease Validated” by Discover Magazine Selected in Top 100 News Stories of 2011.

  1. orthodontics says:

    I love your blog.. very nice colors & theme. Did you create
    this website yourself or did you hire someone to do it for you?
    Plz respond as I’m looking to construct my own blog and would like to know where u got this from. thanks

Leave a Reply

Your email address will not be published. Required fields are marked *