• SEREX;
  • neuroblastoma;
  • tumor antigen;
  • tumor marker;
  • Hu syndrome


  1. Top of page
  2. Abstract
  6. Acknowledgements

Autologous serological screening of a cDNA expression library (SEREX) derived from childhood neuroblastoma led to the identification of 10 different antigens, including 6 novel gene products. The novel antigen 018INX was derived from a small open reading frame in a region of α-internexin mRNA that was previously described as 3′ untranslated region. 018INX thus represents a novel type of tumor antigen. Five novel gene products were derived from NNP-1 (NNP3) and Hu genes (HuC-L, HuD3, HuDY, HuD1proc). As indicated by sequence analysis, these antigens were generated by alternative splicing and/or alternative promoter usage or allelic polymorphism. mRNA expression analyses revealed different tissue restrictions of novel compared to known HuD and NNP-1 transcripts in normal and malignant tissues. The expressions patterns of distinct transcripts indicated potential clinical meanings as diagnostic and/or prognostic tissue markers. When kinetics of serum antibody titres against SEREX-defined antigens were compared to tumor load over time in our patient with neuroblastoma, we found 100-fold increases of anti-Hu and anti-018INX antibody titres preceding the clinical diagnosis of recurrent tumor growth after 2 years. When sera of pediatric patients with cancer (30) and healthy controls (30) were tested for humoral responses to SEREX-defined neuroblastoma antigens, we detected antibodies against all known antigens and NNP3 with low frequencies and titres in control sera, while anti-018INX and anti-Hu antibodies were found in cancer patients only. Our findings indicate that SEREX-defined tumor antigens might provide novel tools for understanding and treatment of this aggressive childhood malignancy. © 2002 Wiley-Liss, Inc.

Neuroblastoma is an embryonal cancer of the sympathetic nervous system and the most common extracranial solid tumor in children. It accounts for 8–10% of all pediatric malignancies and for approximately 15% of cancer-related mortality in children. The tumour's clinical features are remarkably heterogenous, including spontaneous tumor regression and maturation as well as aggressive tumor growth. Neuroblastoma is frequently advanced at the time of diagnosis and associated with poor prognosis, calling for novel diagnostic and prognostic markers, a better understanding of tumor biology and novel therapeutic concepts.1.

The long-standing appreciation of neuroblastoma as an immunogenic tumor2 has recently gained new clinical attention. Antineuronal antibodies in patients with neuroblastoma have been correlated with a more favourable clinical outcome.3, 4 Immunotherapeutic approaches with antibodies5, 6 or tumor cell vaccines7 are currently under investigation and tumor antigens identified in various cancer types of adults are re-evaluated for their clinical use in the context of neuroblastoma.8, 9, 10 Recently, tubulin isoforms were identified as tumor antigens in protein extracts of neuroblastoma,11 suggesting that additional tumor antigens are expressed in this pediatric tumor.

To fully explore the antigen repertoire of neuroblastoma for clinical options, further screening for tumor antigens seems promising. Numerous tumor antigens in cancers of adults have been identified by tumor-specific Tcells.12, 13 or serum antibodies.14, 15 These antigens include viral proteins, proteins overexpressed, aberrantly expressed or molecularly alterated in cancers and cancer-specific products of alternative translation. Many of them contributed to novel diagnostic concepts, to the understanding of tumor development and/or to antigen-specific immunotherapy.12, 13, 14, 15, 16, 17

The large number of interesting tumor antigens identified by SEREX14, 15, 18 prompted us to apply this method for the first time to neuroblastoma tissue. We identified 10 different tumor antigens and provide genetic as well as serological evidence for their possible clinical implications.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Construction of a CDNA Expression Library

The study had been approved by the ethical review board of the Technical University of Munich. Tumor tissue of a 31-month-old girl with neuroblastoma stage 3 (INSS staging system),19 grade 2 (Hughes' grading system)20 without MYCN amplification or chromosome 1 alteration21 was obtained as surgical specimen, snap frozen in liquid nitrogen and stored at −80°C. Poly (A)+RNA was isolated from mechanically homogenised frozen tissue by using an mRNA isolation kit (Stratagene, La Jolla, CA). The cDNA expression library was constructed, packaged and used to infect Escherichia coli (E.coli) XL-1 Blue MRF′ (XL-1) according to instructions to the λZAPII cDNA Gold cloning kit (Stratagene, La Jolla, CA), resulting in more than 1×106 primary recombinants.

Preadsorption of Sera

Sera from 30 pediatric cancer patients including the 31 months old girl with neuroblastoma were obtained after informed consent of their parents during diagnostic or therapeutic procedures and stored at −80°C, together with sera from 30 healthy volunteers. To remove antibodies reactive with antigens related to the vector system, each serum was preadsorbed to XL-1 and to non-recombinant λZAPII phage particles by the following procedure: Preadsorption columns were prepared by using glutaraldehyde-activated affinity adsorbens (Roche Diagnostics, Mannheim, Germany) and XL-1 either sonicated or lysed by λ-phage infection. Preadsorption nitrocellulose membranes were prepared as plaque lifts from XL-1 cultures lytically infected with λ-phages and grown on Luria-Bertani (LB) agar plates. For preadsorption, 2 ml of serum were diluted 1:10 with Tris-buffered saline (TBS), incubated subsequently with column bound proteins of sonicated and lysed XL-1 (over night, 4°C) and then with 6 plaque lifts (3 hrs each, room temperature). Sera were then further diluted 1:10 in TBS (0.2% low-fat dried milk), stabilised by addition of 0.0125% sodium acide and thimerosale and stored at 4°C.

Immunoscreening of CDNA Libraries

For library screening, a total of 1×106 recombinant clones was tested as described by Sahin et al.,18 with the modification that serum was primarily diluted only 1:100 as reported by Scanlan et al.22 Briefly, λ-phage infected XL-1 were plated on LB-agar (3.5 × 103 plaque forming units per 15 cm plate) and expression of recombinant proteins induced with isopropyl β-D-thiogalactoside (IPTG). Following overnight incubation, proteins were transferred to nitrocellulose membranes, which were washed with TBS-T (TBS plus 0,5% Tween 20), blocked with 5 % low-fat dried milk in TBS and then incubated with preadsorbed serum over night at room temperature. Next day, membranes were incubated with an alkaline phosphatase-conjugated anti-human IgG secondary antibody (Dianova, Hamburg, Germany) and stained with 4-nitroblue-tetrazolium-chloride/5-bromo-4-chloro-3-indolyl-phosphate. Positive clones were subcloned to monoclonality by using the same plaque assay procedure. Recombinants expressing IgG from tumor-infiltrating B-cells were identified by staining membranes without prior serum exposure. To determine the serum titre of antibodies reactive with recombinant clones, the serum was tested in plaque assays at 10-fold serial dilutions ranging from 1:100 to 1:100,000.

Sequence Analysis of Reactive CDNA Clones

Positive λ phage clones were converted into phagemids (Stratagene, La Jolla, CA) by in vivo excision23 and plasmid DNA purified over Qiagen Maxi Prep columns (Qiagen, Hilden, Germany). After sequencing of the cDNA inserts, GenBank databases ( and the SEREX database ( were screened for homologies. Novel sequences were submitted to the GenBank database. The respective accession numbers are indicated in Table I. Corresponding genomic DNA sequences were analysed using the Internet address (programs: FGENES, FEXH, HSPL, HEXON, TSSG, TSSW and NNP). Open reading frames were predicted by computer programs (CLONE MANAGER 5, GENETOOL).

Table I. Molecular Characterization of Serex-Defined Antigens Cloned From a Childhood Neuroblastoma
Designation (GenBank Access Number)Number of clones isolatedcDNA HomologyBiological significance of proteins encoded by homologous mRNAsNovel mRNA characteristicsNovel protein characteristics
SEREX databaseGenBank database
HuDpro1 (AY033998)1 M62843.1 (HuDpro) XM_018255.1 (ELAVL4)Neuron-specific mRNA binding proteinExtended 5′UTRNone
HuDpro1c1 Single bp substitutionSingle aa substitution
HuD1 (AY033997)5 M62843.1 without bp 868-909 (HuDpro splice variant HuD)Extended 5′UTRNone
HuD3 (AY033995)1 Alternative 5′ UTR/5′ CDSAlternative N terminus
HuD4 (AY033996)2 Alternative 5′ UTR/5′ CDSAlternative N terminus
HuC-L (AY034002)1 L26405.1 (HuC) D26158.1 (PLE21) NM_001420.2 (ELAVL3)Insertion of 21 bpInsertion of 7 aa
NNP3 (AY033999)1NGO-St-143U79775.1 (NNP-1/Nop52) AL137757.1 (DKFZp434C085)Nucleolar protein involved in pre-rRNA processingAlternative 5′ UTR/5′ CDSAlternative N terminus
018INX (AY034000)2 NM_032727.1 (α-internexin/NF66)Neuron-specific intermediate filamentOnly 3′UTR clonedSecond downstream gene product
018NAC (AY034001)1NY-Co-2AF054187.1 (αNAC) X80909.1 (NACA)Transcriptional coactivatorNoneNone
018HSP1NY-REN-38NM_005348.1 (Hsp90α)Molecular chaperoneNoneNone

Serological Analysis of Positive Clones

Sequential autologous serum samples were obtained at the time of initial diagnosis and at 8, 10, 20, 24 and 27 months after initial diagnosis. Within the first 8 months, the patient received incomplete gross tumor resection and chemoradiotherapy as suggested in the german protocol NB 97.1 The last serum sample was obtained when recurrent tumor growth was suspected from rising NSE and LDH levels and documented sonographically. The histopathological examination of biopsy specimen obtained at the time of initial diagnosis, 5 and 28 months after initial diagnosis revealed neuroblastoma Hughes' grade 2, grade 1–2 and grade 2–3, respectively.

Additional sera were obtained from 29 unrelated children with newly diagnosed neuroblastomas (n = 9), other neuroectodermal malignancies (n = 10) or non-neuroectodermal cancer (n = 10) and from 30 healthy young adults. Reactivities of all sera were tested in plaque assays at 10-fold serial dilutions ranging from 1:100 to 1:100,000.

Analysis of Tissue MRNA Expression by Reverse Transcription and Polymerase Chain Reaction (RTPCR)

Tissue RNA was purchased from Clontech (Palo Alto, CA) (brain, thymus, spleen, liver, pancreas, small intestine, adrenal gland, kidney, prostate and testis) or prepared from surgical or autopsy specimen (tumors, tonsil, skin, lymph node, heart, bladder, uterus, skeletal muscle and ovary) or from pooled peripheral blood mononuclear cells using TRIZOL (Life Technologies, Karlsruhe, Germany). The integrity of mRNA was tested by gel electrophoresis. cDNA was prepared from RNA by using MuLV reverse transcriptase according to the guidelines of the manufacturer (New England Biolabs, Frankfurt, Germany). Each PCR reaction was set up in a total reaction volume of 50 μl using 3 μl of RT product and 1 U of Taq DNA polymerase (Promega, Heidelberg, Germany). After initial denaturation at 95°C for 5 min, 40 cycles were performed. Each cycle consisted of a denaturation step at 95°C (1 min), followed by a primer annealing step (1 min) at the annealing temperature (AT) indicated below and an extension step at 72°C (2 min). Seven primer pairs were designed to differentially amplify 5′ sequences specific for HuD1 or NNP1 cDNAs or their 5′ altered homologues: Forward primers were HuD1se: TCCTAGAATCGGGGGTTTCA (AT: 54°C) [AY033997.1, bp31–50], HuD2se: GTTGACCTGAAGCCAAGAAG (AT: 55°C) [S73887.1, bp27–46], HuD3se: CTGTTGCACGTGAATGCTCT (AT: 58°C) [AY033995.1, bp21–41], HuD4se: GGGGCTGACTGATATGAGAT (AT: 52°C) [AY033996.1, bp30–49], NNP1se: ATGGTTTCGCGCGTGCAGCT (AT: 56°C) [U79775.1, bp39–58], NNP2se: ATCCTCATGTCCAGATCCTG (AT: 55°C) [AL137757.1, bp216–235] and NNP3se: GCTGCCTGTCATGTTTGCTT (AT: 54°C) [AY033999.1, bp29–48]. Reverse primers were HuDan: GTAGACAAAGATGCACCACC [M62843.1, bp1000–981] and NNPan: ACATCACTCCCTGCGTTTCT [U79775.1, bp 1427–1408]. As a control, the published primer pair 3′HuDse: CCAGGCCCTGCTCTCCC and 3′HuDan: AGGCTTGTCATTCCATC (AT: 55°C) [M62843.1, bp748–764 and bp945–929] was used to amplify downstream sequences of HuD cDNA common to all HuD cDNA variants.24 HuC cDNA sequences were amplified using another published pair of primers: HuCse: 5′-AACAACCCAAGTCAGAAGAC and HuCan: 5′-TTGTACACGAAGATGCACCA (AT: 50°C) [bp756–775 and bp991–972, L26405.1].24 After size-separation by gel electrophoresis (1% agarose in TAE buffer), PCR products were transferred to Hybond N+ Nylon membranes (Amersham Pharmacia, Freiburg, Germany), UV-crosslinked and hybridised with 32P-labelled probes. All PCR reactions and subsequent Southern blot hybridisations were performed in at least duplicates.

Southern Blot Hybridisation of PCR Products

To prove identity of PCR products and to increase sensitivity of mRNA analysis, Southern blot hybridisation of HuD-, HuC- and NNP1-related PCR products was performed using radioactively labelled specific internal oligonucleotides. Membranes were hybridised for 15 hrs in Church's buffer25 at 65°C with the following polynucleotide kinase-labelled (γ32P-dATP) oligonucleotids: HuD1: TATTGGGTCTCGCAGAGCTTCGACTCTTCT [AY033997.1, bp244–215], HuD2: CATGTAAGTAATTTAAGTGGCTCCACTTCT [S73887.1, bp93–64], HuD3: ATGGTGCTAATTATCTGTTCCATATGACTG [AY033995.1, bp105–76], HuD4: GCTAATTATCATAAATGAGCAGTTTCTTGACTC [AY033996.1, bp111–79], 3′ HuD: GTGGTGAAGTGGACCTGGGTAGCGCCGGTT [M62843.1, bp 811–782] which recognises all 3 HuD 3′ splice variants described,26 NNP1: CTCGATCTGTCTTTCTTCCCAGCCTTGCATCTTCAGAACCTTCAAGGACTCG TTCAGGAC [U79775.1, bp470–411], NNP2: CCAGGCCCTACCCCCAGAACTGGATCTT GC [AL137757.1, bp270–241], NNP3: GCTCCTCGATCTGTCTGAGGAAAGAAAAAGC [AY033999.1, bp75–45] and HuC-L: GGCGATGAGCGACAGGGGACT (AY034002.1, bp740–720). Membranes were washed with 2 × SSC/1 % SDS and exposed to Kodak XR films for up to 7 days using intensifying screens at −80°C.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Serological Identification and Sequence Analysis of Tumor-associated Antigens

To identify tumor antigens, a neuroblastoma cDNA expression library was screened with autologous serum taken at the time of diagnosis and diluted 1:100. Of more than 1×106 clones tested, 16 were recognised by serum IgG antibodies, coding for 10 different antigens (Table I).

Six antigens belong to the family of Hu proteins. This protein family had originally been defined by high titred serum antibodies in a patient with small cell lung cancer (SCLC).27 These patients suffered from paraneoplastic autoimmune neuropathies summarised as Hu syndrome.28, 29 Hu proteins are RNA binding proteins highly homologous to the Drosophila protein ELAV (embryonic lethal abnormal vision)30 and characterised by 2 tandem RNA recognition motifs (RRMs), an alternatively spliced basic domain and a third RRM.26

The SEREX-defined antigens HuDpro1 and HuD1 are identical to HuDpro (M62843.1) and its splice variant HuD which lacks 14 amino acids of the basic protein domain.26 The 5′ untranslated region of HuDpro1 and HuD1 cDNAs extend beyond the published HuDpro/HuD 5′ untranslated region and correspond to genomic sequences located on chromosome 1 (AL583843.3, bp 6939–7204) right upstream of published HuD sequences (Fig. 1a). Further upstream, a putative TATA box and promoter (AL583843.3, bp 6906 and 6937) were identified by computer-assisted motif searches (Fig. 1a), suggesting that transcription of HuDpro1 and HuD1 is driven by this promoter.

thumbnail image

Figure 1. Schematic 5′ exon maps and N-terminal amino acid sequences of known and novel HuD (a) and NNP-1/Nop52 (b) variants. The 5′ sequences of HuD1 (M62843.1)(26), HuD2 (S73887)(32), NNP-1 (U79775.1)(33) and NNP2 (AL137757.1) were described earlier. The relative positions of novel (filled boxes) and known (hatched boxes) exon sequences within the HuD and the NNP-1 gene are given for all 5′ variants. Positions referring to GenBank entries AL583843.3 (a) and AC003656.1 (b) are indicated as mentioned in the text. The predicted intron promoters at which transcription of HuD1 and NNP3 mRNA might be initiated are marked (bent arrows). A potential splice acceptor site in the HuD gene (full triangle) might be relevant to the generation of HuD2 mRNA. A splice acceptor site relevant to the generation of HuD1-4 (open triangle) was suggested earlier by Sekido et al.32 The location of specific upstream primers (short bold lines above exon boxes) used in PCR experiments and predicted translation initiation sites (ATG) within cloned cDNAs are indicated. N-terminal novel (bold letters) and known (standard letters) amino acid sequences are shown as far as they differ from each other. In HuD4, NNP1 and NNP3 mRNAs, predicted start codons are not preceded by a stop codon and therefore, the respective proteins possibly have an extended N-terminus as indicated by (…) in HuD4 and NNP1 and by (MGE…LPV) in NNP3. This putative 17aa amino acid sequence of NNP3 resulted from translation of genomic sequences downstream of the indicated ATG (AC003656.1, bp 468519) in intron 5 of the NNP-1 gene.

Download figure to PowerPoint

The antigen HuDpro1c differs from HuDpro1 by a single amino acid. A substitution of thymidine by cytosine (M62843.1, bp 902) leads to a serine/proline exchange within the HuDpro-specific 14 amino acids of the basic domain. As with others,31, 32 we did not find any evidence of tumor specific mutation. Instead, allelic polymorphism could be confirmed at this site by amplification and sequencing of the original tumor cDNA and genomic DNA from the patient's tumor and B cells (data not shown).

The antigens HuD3 and HuD4 are homologous to HuD1 but carry alternative amino termini (Fig. 1a). The 5′ ends of the respective cDNAs (AY033995.1, bp 1-91 and AY033996.1, bp 1-102) correspond to genomic sequences on chromosome 1 (AL583843.3, bp 1916–2006 and 4298–4399), which are located upstream of the above described HuD promoter, indicating transcription from additional and yet unidentified upstream HuD promoter(s). The novel exon sequences are spliced to the HuDpro coding sequence at a position (M62843.1, bp 104) where alternative splicing has been described earlier,32 possibly leading to translation of a truncated HuD variant, named here HuD2 (Fig. 1a). Taken together, these results suggest that alternative splicing as well as alternative promoter usage might account for the heterogeneity of HuD gene products.

A 6th Hu antigen is highly homologous to a long murine HuC splice variant (U29148.1) and was thus designated HuC-L. The known human homologues, HuC (L26405.1), PLE21 (D26158.1) and ELAVL3 (NM_001420.2), lack the amino acid sequence SPLSLIA present in the basic region of the SEREX-defined antigen and in murine HUC-L (Fig. 2). The corresponding sequence in the cDNA clone (AY034002.1, bp 720–740) is identical to genomic sequences on chromosome 19 (AC008481.1, bp 103767–103787) and most likely constitutes an alternatively spliced exon of the human ELAVL3/HuC gene.

thumbnail image

Figure 2. Sequence comparison of HuC-L homologues. Regions of sequence variation are indicated by hatched boxes. Differences from HuC-L are underlined in murine (m) HuC-L (U29148.1), human HuC (L26405.1), human PLE21(D26158.1) and human ELAVL3 (NM_001420.2). HuC-L and its murine homologue differ by only 2 amino acids (position 109 and 112 in HUC-L).

Download figure to PowerPoint

The antigen NNP3 was identified as novel amino terminal variant of the nucleolar autoantigen NNP-1/Nop52,33, 34 named here as NNP1. NNP3 adds to another amino terminal variant putatively translated from a testis cDNA (AL137757.1), designated here as NNP2 (Fig 1b). NNP-1/Nop52 is the human homologue of the yeast ribosomal RNA processing protein RRP1 and is most likely involved in the generation of 28S rRNA.34 The NNP-1 gene is located on chromosome 21 in close proximity to several candidate disease genes.33 The 5′ end of the NNP3 cDNA and homologous EST sequences (BE720264) correspond to genomic sequences within intron 533 of the NNP-1 gene (AC003656.1, bp 468532-592 and bp 468230-592). Two putative TATA boxes (AC003656.1, bp 466689 and 468039) and promoters (AC003656.1, bp 466719 and 468082) were identified in intron 5 by computer-assisted motif searches (Fig 1b), suggesting that transcription of NNP3 is initiated at either one or both of these sites. The predicted coding sequence begins at an AUG that corresponds to the ATG indicated in Figure 1b (AC003656.1, bp 468519).

018INX was identified as novel gene product of the 3.2 kb NF66/α-internexin mRNA (NM_032727.1), which is known to code for a 66 kDa neurofilament predominantly expressed in the developing brain and in 50% of neuroblastoma.35, 36 The SEREX-defined antigen was encoded by a region of NF66/α-internexin mRNA that was formerly recognised as part of the 3′ untranslated region. Computer-assisted sequence analyses predicted 6 different open reading frames (bp 19–66, 47–100, 165–407, 478–594, 494–571 and 594–650, AY034000.1), the largest of which (bp 165–407) is depicted in Figure 3.

thumbnail image

Figure 3. 018INX cDNA and largest predicted amino acid sequence. Position 1 in the 018INX cDNA corresponds to position 2481 in NF66/α-internexin cDNA (NM032727.1). Differences from NF66/α-internexin cDNA (NM032727.1) are underlined. Computer-assisted sequence analyses predicted 6 different open reading frames (bp 19–66, 47–100, 165–407, 478–594, 494–571 and 594–650). Translation from the largest open reading frame (bp 165–407) resulted in a 81 aa sequence.

Download figure to PowerPoint

The 018NAC antigen is identical to the ubiquitous nascing-polypeptide-accociated complex protein alpha (α-NAC/NACA), which is recognised as transcriptional coactivator.37 Although 018NAC cDNA sequences slightly differed from the α-NAC and NACA cDNA sequences (AF054187.1, X80909.1), it exhibited 100% homology to a SEREX-defined cDNA cloned from colon cancer (NY-Co-2) (SEREX database).

The antigen 018HSP was formerly identified as NY-REN-38 in renal cancer (SEREX database)22 and is identical to Hsp90α (NM_005348.1), a stress inducible molecular chaperone that is predominantly expressed in brain and testis and overexpressed in many cancers.38

Correlation of Antibody Responses Against SEREX-defined Antigens with Clinical Events

To assess possible changes in humoral immune responses against SEREX-defined antigens over time, 6 sequential serum samples were collected from the patient over a period of 27 months (Table II). To quantify antibody responses, 10-fold serial dilutions of serum samples were tested for recognition of antigens in plaque assays. Using serum taken at the time of initial diagnosis, 018NAC was recognised by dilutions of maximally 1:100, NNP3, 018INX and 018HSP by dilutions of up to 1:1,000 and Hu antigens by dilutions of up to 1:10,000. Concomitant with induction of partial clinical remission and tumor maturation, antibody titres dropped and reached lowest levels 8 to 20 months after initial diagnosis, when humoral responses to 018INX, 018NAC and 018HSP were no longer detectable. Between 20 and 24 months after diagnosis, antibody titres rose up to 100-fold for Hu antigens, 018INX and 018NAC, while neither the tumor volume determined by sonography nor serum levels of tumor markers or total IgG changed significantly. Anti-018INX antibody titres reached maximum levels at 27 months after diagnosis, when disease progression was finally diagnosed by a 5-fold increase in tumor volume and 1.6- to 4-fold increases in serum levels of tumor markers. Remarkably, serum titres of anti-018INX and anti-018NAC antibodies were 10-fold higher at the time of recurrent tumor growth than at initial diagnosis, which might reflect a more malignant phenotype.

Table II. Titer Kinetics of Autologous Serum Antibodies Against Serex-Defined Neuroblastoma Antigens Over Time in the Course of Disease1
Time interval (months)4Tumor volume (ml)8NSE/serum (μg/l)9LDH/serum(U/l)10total IgG/serum (mg/dl)11SEREX-defined antigens
Hu antigens018INX018NAC018HSPNNP3
  • 1

    Sequential autologous serum samples were taken from the young patient with neuroblastoma at the indicated time intervals after initial diagnosis and tested in plaque assays at dilutions ranging from 1:100 to 1:100,000.

  • 2

    Maximal dilutions resulting in antigen recognition are indicated.

  • 3

    No antigen recognition.

  • 4

    Months after initial diagnosis.

  • 5

    At this time abdominal neuroblastoma was diagnosed (Hughes' grade 2) and incompletely resected.

  • 6

    Within 8 months after initial diagnosis the patient received chemotherapy, a second subtotal tumor resection (Hughes' grade 1–2) and radiotherapy.

  • 7

    28 months after initial diagnosis the patient received a third subtotal tumor resection (Hughes' grade 2–3).

  • 8

    Tumor volumes were determined by sonography.

  • 9

    Normal range of neuron specific enolase (NSE) serum levels: < 15 μg/l.

  • 10

    Normal range of lactatdehydrogenase (LDH) serum concentration: < 300 U/l.

  • 11

    Normal levels of total IgG serum levels: 500–1,300 mg/dl.


Recognition of SEREX-defined Antigens by Sera of Cancer Patients and Healthy Individuals

To determine frequency, specificity and titres of antibody responses against the SEREX-defined antigens, sera from 30 pediatric patients with malignant disease including our young neuroblastoma patient and sera from 30 healthy controls were tested in the plaque assay (Table III). Antibodies against all antigens occurred more frequently in pediatric patients with cancer than in healthy controls (Table IIIA) and all but anti-018HSP antibodies occurred at higher titres in the context of cancer (Table IIIB). Frequencies of anti-018HSP and anti-Hu antibodies were high in neuroblastoma patients (5/10 and 2/10), while anti-018NAC antibodies occurred most frequently in patients with Ewing's sarcoma (2/3) and non-Hodgkin's lymphoma (2/2). Humoral responses against Hu proteins and 018INX were found in sera of cancer patients only (5/30 and 1/30), whereas antibodies against 018HSP, 018NAC and NNP3 occurred in patients with cancer as well as in some healthy individuals (9/30, 6/30 and 2/30 compared to 7/30, 3/30 and 1/30, respectively). The induction of humoral immune responses against the latter antigens may thus occur independently of malignant disease.

Table III. Frequency (a) and Titers (b) of Serum Antibodies Against Serex Defined Neuroblastoma Antigens in Pediatric Patients With Cancer and Healthy Volunteers1
AHu antigens018INX018NAC018HSPNNP3
  • 1

    Sera from 30 healthy volunteers, from our patient with neuroblastoma and from 29 unrelated pediatric patients with cancer were tested in plaque assays (dilution 1:100–100,000).

  • 2

    Numbers in parentheses indicate numbers of individuals investigated.

  • 3

    n = individuals with serum antibodies against the SEREX-defined neuroblastoma antigens.

  • 4

    Maximal dilutions that result in antigen recognition.

  • 5

    n = individuals the serum of whom stained SEREX-defined neuroblastoma antigens at the indicated dilution.

Healthy volunteers (30)2030371
Patients with cancer (30)51692
 Neuroblastoma (10)21151
  Stage 1 (2)00010
  Stage 3 (3)21131
  Stage 4s (2)00000
  Stage 4 (2)00010
  Relapsed stage 4 (1)00000
 Other neuroectodermal malignancies (10)20210
  Medulloblastoma (4)00010
  Astrocytoma (3)10000
  Ewing's sarcoma (3)10200
 Non-neuroectodermal malignancies (10)10331
  Rhabdomyosarcoma (2)00010
  Osteosarcoma (2)00010
  Hodgkin's disease (2)00001
  Non Hodgkin's lymphoma (2)10210
  Acute lymphoblastic leukemia (2)00100
BHu antigens018INX018NAC018HSPNNP3
Healthy volunteers (30)00371
Patients with cancer (30)51692

Expression of Novel Hu and NNP-1 MRNA Variants in Normal and Malignant Tissues

To determine mRNA expression patterns of the novel variants of HuD, HuC-L and NNP-1, we performed RT-PCR using 23 normal tissues (Table IVA) and 18 pediatric tumor specimen (Table IVB). To achieve maximal specificity and sensitivity, RT-PCR was followed by Southern blot hybridisation by using radioactively labelled gene-specific internal oligo-nucleotides. Of the different Hu mRNA variants tested, HuD1 transcripts displayed the most restricted expression pattern. HuD1 mRNA was not detected in any of the normal tissues except brain and exclusively in tumors of neuroectodermal origin. By contrast, HuD 2, 3 and 4 as well as HuC-L mRNA was found in brain as well as in some extracranial normal tissues and in the majority of neuroectodermal tumors including all neuroblastomas. Moreover, HuD2, HuD3 and HuC-L transcripts occurred in some non-neuroectodermal malignancies. NNP2 and NNP3 mRNA expression was restricted to distinct normal and malignant tissues, while NNP1 mRNA was expressed ubiquitously as described.33 The data obtained from a small number of cancer tissues show that HuD1 and NNP3 mRNA expression is lost in some neuroblastomas associated with aggressive disease progression. Furthermore, downregulation of NNP3 mRNA was associated with MYCN gene amplification recognised as reliable marker for poor prognosis of neuroblastoma.1, 21 NNP3 mRNA expression was also undetectable in 2/4 medulloblastomas and 1/2 osteosarcomas associated with therapy-resistant disease progression or death due to disease (Table IVB).

Table IV. MRNA Expression of Hud and NNP1 5′Variants and HUC-l in Normal (a) And Malignant (b) Tissues1
  • 1

    RT-PCR and Southern blot hybridization.

  • 2

    PBMC, peripheral blood mononuclear cells.

  • 3

    NB, neuroblastoma; MB, medulloblastoma; AC, astrocytoma; EP, ependymoma; OS osteosarcoma; RMS, rhabdomyosarcoma; HD, Hodgkin's disease; NHL, non-Hodgkin's lymphoma.

  • 4

    INSS stage,

  • 5

    CR, complete remission; PR, partial remission; PD, progressive disease; D, death due to disease;

  • 6

    month after diagnosis.

  • 7

    a, amplified MYCN gene; n, non-amplified MYCN gene.

Brain   ++++++++
Thymus   ++++
Tonsil   +++
Lymph node   +++
Spleen   ++
Pooled PBMC2   +++
Heart   +++
Lung   ++
Thyroid   +
Liver   ++++
Pancreas   ++++++
Small intestine   +++++++
Bladder   ++++
Adrenal gland   +++++
Kidney   ++++
Prostate   +++++++
Testis   ++++++
Breast   ++
Uterus   +++
Ovary   ++
Placenta   ++
Sceletal muscle   ++
Skin   ++
Neuroblastomas (6)
NB 3PR37n++++++++
NB 3PR31n++++++++
NB 4sCR57n++++++++
NB 4Da++++++
NB 4Da++++++
Other neuroectodermal malignancies (6)
MBCR24 +++
MBCR47 ++++++
MBPD37 ++++++
MBD ++++++
ACCR32 +++++++
EPCR21 ++++++
Non-neuroectodermal malignancies (6)
OSCR24 +++
OSD +++
RMSCR27 +++++
HDCR75 ++++
HDCR73 ++++
NHLCR18 +++


  1. Top of page
  2. Abstract
  6. Acknowledgements

In our study, we examined the antitumoural immune response in a young child with neuroblastoma. By applying for the first time the SEREX methodology to a pediatric tumor, 10 tumor antigens were identified and their potential clinical implications assessed. Four SEREX-defined antigens were identical to Hsp90α, αNAC, HuD and HuDpro and 6 were novel products of the HuD, HuC, NNP-1 and α-internexin genes.

The young patient was followed for changes in the respective antibody titres during initial response to therapy, partial clinical remission and subsequent disease progression up to 27 months after initial diagnosis. We observed an upregulation of humoral immune responses against Hu antigens, 018INX and 018NAC prior to clinically evident disease progression. The increase of specific antibody titres was not due to recovery from immune suppression as indicated by permanent low total IgG serum levels. It might rather be attributed to a possible increase of antigen expression with histological progression to the more malignant phenotype seen at 28 months after initial diagnosis. Unfortunately, no tumor tissue was available for comparative antigen expression analyses. However, our findings corroborate earlier observations that correlate clinical events with changes in humoral immune responses to a SEREX-defined antigen,39 and suggest that the specific antibody responses may be useful for disease monitoring.

Antibodies against 018INX and, as expected from the literature,28 antibodies against Hu antigens were found to occur only in the context of cancer and thereby filled the prerequisite as diagnostic tumor markers. Anti-Hu antibody frequencies in our patients with neuroectodermal cancers (20%) (2/10) were in the same range as reported for adult SCLC patients (16%) (32/196).40 In children with neuroblastoma, we observed somewhat higher anti-Hu antibody frequencies (20%) (2/10) than reported by others (4%, 16%) (4/109, 10/64).4, 41 This was most likely due to higher test sensitivity in our study achieved by lower serum dilutions (1:100 vs. 1:250–500 in the other studies), though differences due to varying patient cohorts can not be excluded. The frequency of anti-018INX antibodies in pediatric patients with neuroblastoma (1/10) should be re-evaluated in larger cohorts.

None of our 2 anti-Hu positive neuroblastoma patients suffered from paraneoplastic neuropathy. This was in line with observations by Dalmau et al.4 and Antunes et al.41 who reported about 1/4 and 6/10 anti-Hu positive neuroblastoma patients without paraneoplastic disease, respectively. However, anti-Hu antibody titres of 1:10,000 in our study were remarkably high, considering the findings of Graus et al.28 who observed neurological symptoms in almost all adult cancer patients with anti-Hu antibody titres of more than 1:1,000. Taken together, anti-Hu antibodies should be recognised as tumor markers independently of any paraneoplastic disease in adults and children and for this purpose warrants investigation at low serum dilutions.

Our results further indicate that anti-Hu antibodies in children are not restricted to patients with cancer of neuroectodermal origin but can also occur in the context of hematological malignancies. This finding adds to earlier reports describing non-classical neuroectodermal tumors associated with humoral anti-Hu immunity.28, 42 However, since tumor tissue from the anti-Hu positive patient with B cell lymphoma was not available, the postulated tissue expression of Hu antigens could not be examined.

Anti-018HSP, anti-018NAC and anti-NNP3 antibodies were detected more frequently (9/30, 6/30 and 2/30) though not exclusively in cancer patients (7/30, 3/30 and 1/30 positive healthy controls). Since prognostic implications of frequent tumor-specific (anti-Hu) as well as non-tumor specific (anti-Hsp90) antibodies are well documented in adult cancer patients,40, 43 we suggest a prospective evaluation of frequent neuroblastoma-associated antibody responses as prognostic markers in childhood malignancies including neuroblastoma.

The identification of novel HuD and NNP-1 gene transcripts may allow new insight into the expression of the respective genes and their clinical relevance. We provide evidence for complex alternative exon and promoter usage that results in alternative amino terminal sequences of HuD and NNP-1 gene products. Our data suggest that the activity of novel promoters located within introns in both genes might be related to the biological behaviour of distinct cancer types. Downregulation of HuD1 mRNA in an aggressively growing neuroblastoma with MYCN gene amplification corroborates earlier correlations of reduced total HuD mRNA levels in aggressive types of neuroblastoma and SCLC.44, 45 Remarkably, downregulation of NNP3 mRNA was not only seen in both neuroblastomas positive for MYCN gene amplification but also in 2 medulloblastomas and an osteosarcoma characterised by aggressive progression. It seems challenging to study possible relations between intron promoter activity in the HuD and NNP-1 gene on chromosome 1p3446 and 21q2233 and known markers of poor prognosis such as MYCN gene amplification on chromosome 2p24 and/or loss of heterozygosity at chromosome 1p36.1, 21 However, a larger number of patients is needed to fully evaluate HuD1 and NNP3 mRNA as possible prognostic tissue markers.

The observation of basically 2 patterns of HuD related mRNA distribution in normal tissues further supports the idea that transcription of the HuD gene is initiated at 2 differentially regulated promoters. HuD2, 3 and 4 mRNAs were detected in some peripheral tissues and most likely account for the faint antigen staining seen in testis, gut and adrenal gland.47, 48 No data are available to decide whether peripheral antigen derives from contaminating neural contents or parenchymal cells. By contrast, HuD1 mRNA, initiated at the putative intron promoter, appears to be restricted to the central nervous system. Whether the respective HuD protein variants fulfil different biological functions or elicit different kinds of immune responses remains to be determined. With respect to the unknown mechanisms of Hu immunogenicity,29 it might be interesting to compare levels of anti-Hu antibody responses in the serum to expression levels of the various HuD transcripts in Hu positive tumors.

The fact that none of the specific 5′ sequences of HuD mRNA variants could be detected in blood or hematopoietic tissues of healthy individuals encourages new attempts to diagnose minimal systemic neuroectodermal cancer by RT-PCR. An earlier approach by means of downstream HuD PCR failed, probably due to amplification of HuD sequences displaying a less tissue restricted expression pattern (reference 49 and our own unpublished observations).

Finally, our study provided first evidence that sequences of the α-internexin mRNA previously described as 3′ untranslated region encode a small gene product that is recognised by the immune system in the context of cancer. These findings were supported by recent Western blot analyses, which demonstrated that the recombinantly expressed protein derived from the largest open reading frame (bp165–407) is recognised by the patient's serum (article in preparation). Usage of alternative open reading frames has been identified before as an important mechanism contributing to the increasing repertoire of T cell epitopes in cancer tissue.13, 50 However, to our knowledge, 018INX is the first example of a tumor antigen translated from a second open reading frame within sequences previously described as 3′ untranslated region. To further evaluate the possible use of 018INX in immunotherapy of neuroblastoma, it is necessary to investigate whether cancer-associated immune recognition is due to cancer-restricted expression or to aberrant expression of this gene product in the periphery.

Taken together, our study shows that the SEREX approach to identification of pediatric tumor antigens is feasible and may help to identify clinical markers as well as targets for the immunotherapy of childhood cancer.


  1. Top of page
  2. Abstract
  6. Acknowledgements

We thank Ugur Sahin, Özlem Türeci and Michael Pfreundschuh for introducing us to the SEREX methodology, Karlheinz Wurster, Franz Prantl, Frank Höpner, Juan-Carlos Lenz, Hans Grundhuber, Dieter Sackerer, Manfred Späth and Claudia Götz for their help with asservation of surgical specimen, Georg W. Bornkamm for critically reviewing the article and the numerous volunteers for donating serum.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  • 1
    Berthold F, Hero B. Neuroblastoma: current drug therapy recommendations as part of the total treatment approach. Drugs 2000;59: 126177.
  • 2
    Hellström KE, Hellström I. Immunity to neuroblastomas and melanomas. Annu Rev Med 1972;23: 1938.
  • 3
    Salmaggi A, Luksch R, Forno MG, Pozzi A, Silvani A, Boiardi A, Gasparini M, Nemni R. Antineuronal antibodies in patients with neuroblastoma: relationships with clinical features. Tumori 1997;83: 9537.
  • 4
    Dalmau J, Graus F, Cheung NK, Rosenblum MK, Ho A, Canete A, Delattre JY, Thompson SJ, Posner JB. Major histocompatibility proteins, anti-Hu antibodies and paraneoplastic encephalomyelitis in neuroblastoma and small cell lung cancer. Cancer 1995;75: 99109.
  • 5
    Ollert MW, David K, Vollmert C, Juhl H, Erttmann R, Bredehorst R, Vogel CW. Mechanisms of in vivo anti-neuroblastoma activity of human natural IgM. Eur J Cancer 1997;33: 19428.
  • 6
    Cheung NK, Kushner BH, Cheung IY, Kramer K, Canete A, Gerald W, Bonilla MA, Finn R, Yeh SJ, Larson SM. Anti-G(D2) antibody treatment of minimal residual stage 4 neuroblastoma diagnosed at more than 1 year of age. J Clin Oncol 1998;16: 305360.
  • 7
    Brenner MK, Heslop H, Krance R, Horowitz M, Strother D, Nuchtern J, Grilley B, Martingano E, Cooper K. Phase I study of chemokine and cytokine gene-modified autologous neuroblastoma cells for treatment of relapsed/refractory neuroblastoma using an adenoviral vector. Hum Gene Ther 2000;11: 147788.
  • 8
    Soling A, Schurr P, Berthold F. Expression and clinical relevance of NY-ESO-1, MAGE-1 and MAGE-3 in neuroblastoma. Anticancer Res 1999;19: 22059.
  • 9
    Cheung IY, Barber D, Cheung NK. Detection of microscopic neuroblastoma in marrow by histology, immunocytology and reverse transcription-PCR of multiple molecular markers. Clin Cancer Res 1998;4: 28015.
  • 10
    Sarkar AK, Nuchtern JG. Lysis of MYCN-amplified neuroblastoma cells by MYCN peptide-specific cytotoxic T lymphocytes. Cancer Res 2000;60: 190813.
  • 11
    Prasannan L, Misek DE, Hinderer R, Michon J, Geiger JD, Hanash SM. Identification of beta-tubulin isoforms as tumor antigens in neuroblastoma. Clin Cancer Res 2000;6: 394956.
  • 12
    Boon T, Coulie PG, Van den Eynde B. Tumor antigens recognized by T cells. Immunol Today 1997;18: 2678.
  • 13
    Wang RF, Rosenberg SA. Human tumor antigens recognized by T lymphocytes: implications for cancer therapy. J Leukoc Biol 1996;60: 296309.
  • 14
    Old LJ, Chen YT. New paths in human cancer serology. J Exp Med 1998;187: 11637.
  • 15
    Pfreundschuh M. Exploitation of the B cell repertoire for the identification of human tumor antigens. Cancer Chemother Pharmacol 2000;46: 37.
  • 16
    Knuth A, Jäger D, Jäger E. Cancer immunotherapy in clinical oncology. Cancer Chemother Pharmacol 2000;46: 4651.
  • 17
    Rosenberg SA. Progress in human tumour immunology and immunotherapy. Nature 2001;411: 3804.
  • 18
    Sahin U, Tureci O, Schmitt H, Cochlovius B, Johannes T, Schmits R, Stenner F, Luo G, Schobert I, Pfreundschuh M. Human neoplasms elicit multiple specific immune responses in the autologous host. Proc Natl Acad Sci USA 1995;92: 118103.
  • 19
    Brodeur GM, Pritchard J, Berthold F, Carlsen NL, Castel V, Castelberry RP, De Bernardi B, Evans AE, Favrot M, Hedborg F, Kanéko M, Kemshead J et al. Revisions of the international criteria for neuroblastoma diagnosis, staging and response to treatment. J Clin Oncol 1993; 11: 146677. Comment in: J Clin Oncol 1994; 12: 1991–3.
  • 20
    Hughes M, Marsden HB, Palmer MK. Histologic patterns of neuroblastoma related to prognosis and clinical staging. Cancer 1974;34: 170611.
  • 21
    Maris JM, Matthay KK. Molecular biology of neuroblastoma. J Clin Oncol 1999;17: 226479.
  • 22
    Scanlan MJ, Gordan JD, Williamson B, Stockert E, Bander NH, Jongeneel V, Gure AO, Jager D, Jager E, Knuth A, Chen YT, Old LJ. Antigens recognized by autologous antibody in patients with renal-cell carcinoma. Int J Cancer 1999;83: 45664.
  • 23
    Short JM, Fernandez JM, Sorge JA, Huse WD. Lambda ZAP: a bacteriophage lambda expression vector with in vivo excision properties. Nucleic Acids Res 1988;16: 7583600.
  • 24
    Manley GT, Smitt PS, Dalmau J, Posner JB. Hu antigens: reactivity with Hu antibodies, tumor expression and major immunogenic sites. Ann Neurol 1995;38: 10210.
  • 25
    Church GM, Gilbert W. Genomic sequencing. Proc Natl Acad Sci USA 1984;81: 19915.
  • 26
    Szabo A, Dalmau J, Manley G, Rosenfeld M, Wong E, Henson J, Postler JB, Furneaux HM. HuD, a paraneoplastic encephalomyelitis antigen, contains RNA-binding domains and is homologous to Elav and Sex-lethal. Cell 1991;67: 32533.
  • 27
    Graus F, Cordon-Cardo C, Posner JB. Neuronal antinuclear antibody in sensory neuropathy from lung cancer. Neurology 1985;35: 53843.
  • 28
    Graus F, Keime-Guibert F, Rene R, Benyahia B, Ribalta T, Ascaso C, Escaramis G, Delattre JY. Anti-Hu-associated paraneoplastic encephalomyelitis: analysis of 200 patients. Brain 2001;124: 113848.
  • 29
    Voltz RD, Posner JB, Dalmau J, Graus F. Paraneoplastic encephalomyelitis: an update of the effects of the anti-Hu immune response on the nervous system and tumour. J Neurol Neurosurg Psychiatry 1997;63: 1336.
  • 30
    Robinow S, Campos AR, Yao KM, White K. The elav gene product of Drosophila, required in neurons, has 3 RNP consensus motifs. Science 1988;16: 15702.
  • 31
    Carpentier AF, Voltz R, DesChamps T, Posner JB, Dalmau J, Rosenfeld MR. Absence of HuD gene mutations in paraneoplastic small cell lung cancer tissue. Neurology 1998;50: 1919.
  • 32
    Sekido Y, Bader SA, Carbone DP, Johnson BE, Minna JD. Molecular analysis of the HuD gene encoding a paraneoplastic encephalomyelitis antigen in human lung cancer cell lines. Cancer Res 1994;54: 498892.
  • 33
    Jansen E, Meulemans SM, Orlans IC, Van de Ven WJ. The NNP-1 gene (D21S2056E), which encodes a novel nuclear protein, maps in close proximity to the cystatin B gene within the EPM1 and APECED critical region on 21q22.3. Genomics 1997;42: 33641.
  • 34
    Savino TM, Bastos R, Jansen E, Hernandez-Verdun D. The nucleolar antigen Nop52, the human homologue of the yeast ribosomal RNA processing RRP1, is recruited at late stages of nucleologenesis. J Cell Sci 1999;112: 1889900.
  • 35
    Chan SO, Chiu FC. Cloning and developmental expression of human 66 kd neurofilament protein. Brain Res Mol Brain Res 1995;29: 17784.
  • 36
    Foley J, Witte D, Chiu FC, Parysek LM. Expression of the neural intermediate filament proteins peripherin and neurofilament-66/alpha-internexin in neuroblastoma. Lab Invest 1994;71: 1939.
  • 37
    Yotov WV, Moreau A, St-Arnaud R. The alpha chain of the nascent polypeptide-associated complex functions as a transcriptional coactivator. Mol Cell Biol 1998;18: 130311.
  • 38
    Csermely P, Schnaider T, Soti C, Prohaszka Z, Nardai G. The 90-kDa molecular chaperone family: structure, function and clinical applications: a comprehensive review. Pharmacol Ther 1998;79: 12968.
  • 39
    Jäger E, Stockert E, Zidianakis Z, Chen YT, Karbach J, Jäger D, Arand M, Ritter G, Old LJ, Knuth A. Humoral immune responses of cancer patients against “Cancer-Testis” antigen NY-ESO-1: correlation with clinical events. Int J Cancer 1999;84: 50610.
  • 40
    Graus F, Dalmau J, Rene R, Tora M, Malats N, Verschuuren JJ, Cardenal F, Vinolas-N, Garcia del Muro J, Vadell C, Mason WP, Rosell R, et al. Anti-Hu antibodies in patients with small-cell lung cancer: association with complete response to therapy and improved survival. J Clin Oncol 1997;15: 286672.
  • 41
    Antunes NL, Khakoo Y, Matthay KK, Seeger RC, Stram DO, Gerstner E, Abrey LE, Dalmau J. Antineuronal antibodies in patients with neuroblastoma and paraneoplastic opsomyoclonus. J Pediatr Hematol Oncol 2000;22: 31520.
  • 42
    Dalmau J, Graus F, Rosenblum MK, Posner, JB. Anti-Hu-associated paraneoplastic encephalomyelitis/sensory neuropathy: a clinical study of 71 patients. Medicine 1992;71: 5972.
  • 43
    Conroy SE, Sasieni PD, Fentiman I, Latchman DS. Autoantibodies to the 90kDa heat shock protein and poor survival in breast cancer patients. Eur J Cancer 1998;34: 9423.
  • 44
    Ball NS, King PH. Neuron-specific hel-N1 and HuD as novel molecular markers of neuroblastoma: a correlation of HuD messenger RNA levels with favorable prognostic features. Clin Cancer Res 1997;3: 185965.
  • 45
    King PH. Differential expression of the neuroendocrine genes Hel-N1 and HuD in small-cell lung carcinoma: evidence for down-regulation of HuD in the variant phenotype. Int J Cancer 1997;74: 37882.
  • 46
    Muresu R, Baldini A, Gress T, Posner JB, Furneaux HM, Siniscalco M. Mapping of the gene coding for a paraneoplastic encephalomyelitis antigen (HuD) to human chromosome site 1p34. Cytogenet Cell Genet 1994;65: 1778.
  • 47
    Dalmau J, Furneaux HM, Cordon-Cardo C, Posner JB. The expression of the Hu (paraneoplastic encephalomyelitis/sensory neuronopathy) antigen in human normal and tumor tissues. Am J Pathol 1992;141: 8816.
  • 48
    Dalmau J, Furneaux HM, Gralla RJ, Kris MG, Posner JB. Detection of the anti-Hu antibody in the serum of patients with small cell lung cancer: a quantitative western blot analysis. Ann Neurol 1990;27: 54452.
  • 49
    Tora M, Barbera VM, Real FX. Detection of HuD transcripts by means of reverse transcriptase and polymerase chain reaction: implications for the detection of minimal residual disease in patients with small cell lung cancer. Cancer Lett 2000;161: 15764.
  • 50
    Probst-Kepper M, Stroobant V, Kridel R, Gaugler B, Landry C, Brasseur F, Cosyns JP, Weynand B, Boon T, Van Den Eynde BJ. An alternative open reading frame of the human macrophage colony-stimulating factor gene is independently translated and codes for an antigenic peptide of 14 amino acids recognized by tumor-infiltrating CD8 T lymphocytes. J Exp Med 2001;193: 118998.