[CANCER RESEARCH 58, 3480-3485, August 1. 1998]

Meeting Report

Seventh International Workshop on Ataxia-Telangiectasia1

Nancy Uhrhammer, Jacques-Olivier Bay, and Yves-Jean Bignon2

Centre Jean Perrin, 63000 Clermont-Ferrand, France

Abstract

Ataxia-telangiectasia (A-T) is a rare hereditary syndrome involvingcerebellar degeneration, immunodeficiency, cancer risk, and radiosensi-tivity. Since the cloning of the A-T gene, ATM, in 1995, research on this

pleiotropic disease and its molecular basis has expanded tremendously.ATM is a large protein kinase that appears to be one of the primarysensors of DNA strand-break damage. The vast majority of mutations in

ATM result in truncation and destabilization of the protein, but certainmissense and splicing errors have been shown to result in a less severephenotype. A-T hétérozygoteshave been shown to have a slightly in

creased risk of cancer, but their increased in vitro radiosensitivity does not.seem to result in any in vivo sensitivity. ATM does seem to act as a classictumor suppressor gene in T-prolymphocytic leukemia, and LOH at the

ATM locus is a common event in some tumor types, suggesting a generalrole for ATM in cancer. Recent work has shown the interaction of ATMwith proteins involved in cell cycle control, and the direct phosphorylationof some of these interactors by ATM. ATM knockout mice have beencreated by several groups, and recapitulate the immunodeficiency, radiosensitivity, cancer risk, and fertility defects of A-T, although the effect on

the cerebellum is slight. These diverse topics, and their integration into aglobal understanding of A-T, were the basis of the 7th International A-T

Workshop.

Introduction

A-T3 is a recessive hereditary chromosomal instability syndrome

with a 100-250-fold excess of lymphoid malignancies as well as a

more modest excess of epithelial tumors in patients. There is someevidence that epithelial tumors are more frequent in A-T carriers, andLOH including the A-T locus at Ilq23.1 is a frequent event inovarian, breast, and cervical cancer, suggesting that the A-T geneplays an important role in malignancy. The focus of A-T researchtoday is to define the molecular mechanisms by which the A-T gene

product, ATM, acts to recognize chromosomal damage and preventthis damage from becoming lethal or transforming, as well as to betterunderstand the role that mutations in ATM have in causing cancer, andwhether constitutional heterozygosity has any implications for chemotherapy and radiotherapy.

In addition, A-T involves progressive cerebellar degeneration, immunodeficiency, gonadal dysgenesis, ocular and cutaneous telangiec-tasia, and high serum AFP. A-T patients are sensitive to ionizingradiation, and cultured A-T cells fail to execute cell cycle checkpoints

Received 3/3/98; accepted 6/2/98.1This workshop was held in Clermont-Ferrand, France on November 22-24, 1997.

The workshop was organized by Jacques-Olivier Bay, Yves-Jean Bignon. Richard A.Gatti, Martin F. Lavin, Gilbert Lenoir. M. Stephen Meyn, Yosef Shiloh, A. Malcolm R.Taylor, and Anthony Wynshaw-Boris. Financial and logistical support was generouslyprovided by the Association pour la Recherche sur l'Ataxie Télangieclasie, the Centre

Jean Perrin, the A-T Medical Research Foundation, the Institut National de la Santéel dela Recherche Médicale,Groupama Assurances, (he HôtelCoubertin, the Conseil Régionald'Auvergne, the Ligue Nationale Contre le Cancer. DélayeTransports, Alitalia, de

Bussac. Smilhkline Beecham, Bristol-Myers Squibb. Novartis, théConseil GénéralduPuy-de-Dôme, and CréditAgricole Centre France.

2 To whom requests for reprints should be addressed, at Centre Jean Perrin, 58 RueMontalembert BP392. 63000 Clermont-Ferrand. France.

' The abbreviations used are: A-T, ataxia-telangiectasia; AFP, a-fetoprotein; LOH.

loss of heterozygosity: NBS. Nijmegen breakage syndrome; PARP. poly(ADP-ribose)polymerase; RPA. replication protein A; RN. recombination nodule.

immediately after DNA strand-break damage and exhibit excessive

apoptotic cell death. This pleiotropic disease has intrigued and puzzled researchers for many years, a situation that is gradually beingresolved now that the ATM gene is accessible for molecular characterization, and several strains of Atm knockout mice have been created.

The Seventh International Workshop on A-T was held in Clermont-Ferrand, France on November 22-24, 1997, hosted by the Jean PerrinCancer Center and the Association pour la Recherche sur l'Ataxie

Télangiectasie,a French organization of A-T families. This workshophas been held in various cities every 2-3 years since 1980 and has

followed the early characterization of the complex clinical and cellular phenotype, the efforts to clone the A-T gene(s) by both positional

and functional complementation strategies, and now the characterization of the ATM gene and its role in cellular physiology. As more hasbecome known about the abnormal responses of A-T cells to DNAstrand breaks, and also as it has become clear that relatives of A-T

patients had an increased incidence of cancer, researchers from otherfields have become interested in A-T not just as a rare disease, but as

one involving a gene fundamental to genomic stability. Thus, theworkshops have grown in attendance from a handful of scientists atthe first meeting to over 150 researchers and clinicians from 19countries at the Seventh International Workshop on A-T, accompanied by several A-T patients and their families. Since the cloning of

the ATM gene in 1995, additional comprehensive workshops havebeen sponsored by the A-T Children's Fund.

Gatti ( 1) presented a review of the disease from the first descriptionby Syllaba and Henner in 1926 to the present. The long road from thedescription of the clinical syndrome (2) to the complementationstudies that implied the existence of at least four disease genes (3), theunsuccessful attempts to clone the A-T gene(s) by functional complementation, the genetic localization of the gene(s) to 1lq22-23 (4), the

cloning of the single gene responsible (5), and the current state ofATM molecular biology reminded the participants of how much workby so many researchers had already been accomplished. Shiloh,whose laboratory first announced the identification of the A-T gene,

ATM (5), discussed how the characterization of ATM reconciled thepleiotropic clinical phenotype with a defect in a single protein, bringing the study of A-T full circle. ATM is a protein kinase that respondsto DNA strand-break damage by arresting all phases of the cell cycle

while simultaneously preventing cell death. The lack of these checkpoints, combined with excessive sensitivity to the lethal effects ofDNA strand-break damage, together account for the cancer risk,

immune deficiency, gonadal dysgenesis, and characteristic chromosomal instability seen in A-T, although the exact mechanism ofcerebellar degeneration is still unclear. The complementation of A-T

cells in culture by the introduction of recombinant ATM protein andthe phenotype of Atm—/—mice confirm that the correct gene has

been cloned (6-9). Lavin discussed the molecular role of ATM in cellcycle checkpoints after DNA strand-break damage and the defectsfound in A-T cells in the induction of p53 and other checkpointeffectors (10, 11). In keeping with its role as a primary DNA strand-

break sensor, ATM seems to associate dynamically with numerous

3480

MEETING REPORT

proteins in large nuclear and microsomal complexes with the ability tosignal to a variety of other effector molecules. These three introductory talks concerning the disease itself, the gene responsible and itsprotein product, and the dysfunction of A-T cells in response to DNA

strand breaks set the stage for the meeting, where the latest advancesin the molecular genetics, cell physiology, and epidemiology of ATMwere discussed.

Mutations in the ATM Gene

The Sixth International Workshop on A-T took place on the vergeof the cloning of the A-T gene, so it was most appropriate to beginwith a session discussing ATM mutations found in A-T patients. Here,

the discussion of the variable and constant clinical features wasrevisited by Taylor in his review of the phenotypes of A-T patients in

the United Kingdom with a milder clinical course, milder immunedefects, and/or less severe radiation sensitivity in their cultured cells.The less severe disease of a group of patients who shared a commonA-T mutation on at least one chromosome was explained by a "leaky"

splice-site mutation in the ATM gene and the presence of a small

amount of normal ATM (12). Similar findings were reported byShiloh on a group of Italian A-T variants (13). The phenotype of other

variant patients in whom ATM mutations have not been found may bedue to defects in different genes. Taylor also suggested that theexpression of abnormal ATM protein with either missense mutationsor in-frame deletions may be correlated with the risk of leukemia,

lymphoma, and breast cancer, and that the cancer risk associated withthe A-T syndrome may not be spread evenly among the patients.

The German ATM Consortium, Van't Veer, Prudente, and Gatti

each presented data showing that mutations in ATM can be found inevery part of the gene, and that the great majority truncate the proteinand reduce its expression to undetectable levels. Missense mutationswere described, but they do not seem to be common in A-T. The

majority of ATM mutations were also unique to a family, althoughfounder effects were described. For example, the Costa Rican population harbors five founder mutations responsible for —¿�95%of A-T

alíelesas well as private mutations (14). The study of this populationdemonstrated the usefulness of creating rapid assays for a handful ofpopulation-specific mutations, allowing the detection of most of the

mutations in a new sample of the same population to be done quicklyand cheaply. This approach will not be useful for more outbredpopulations, because the large size of the ATM gene and the lack ofmutational hot spots make screening for mutations time-consuming

and difficult. A database of ATM mutations is accessible at http://www.vmmc.org/vmrc/atm.htm.

The degree of immunodeficiency is a variable feature of the A-T

phenotype (for a review, see Ref. 15). Some patients are troubledcontinuously with severe infections, whereas others suffer only mildor undetectable immunodeficiency. One theory to account for thiswide variation among patients was that different mutations may havedifferent effects on immune gene rearrangement or immune cellsurvival, particularly for missense and in-frame deletion mutations,

where an altered protein may be expressed and have partial function.For example, a recently described patient with classical ataxia andradiosensitivity but no immune defect was found to be a compoundhétérozygotefor a missense mutation in the kinase domain and atruncation of nine amino acids at the carboxyl terminus of ATM, andToyoshima et al. (16) hypothesize that this combination of mutationsmay allow some function of ATM in immune cells. The idea that theimmunophenotype can be predicted from the molecular genotype,however, is not supported by any general correlation between genotype and phenotype, because patients with similar mutations may havedifferent immunophenotypes, and patients with dissimilar mutations

may have similar immune function. Thus, it seems that much of thevariation in this part of the A-T phenotype may be due to other genetic

or environmental factors.

ATM and Cancer

The fact that A-T homozygotes have a greatly increased risk of

cancer is clear, but the risk to hétérozygotesis, as yet, unclear.Stoppa-Lyonnet reviewed the literature concerning this field andpointed out that the risk of breast cancer in known A-T hétérozygoteswas real but was less than previously estimated (e.g., Refs. 17-20).

Lenoir considered which kind of breast cancer families, if any, mightbe due to ATM mutations and suggested that the low penetrance ofATM mutations and their association with later-onset breast cancer inA-T families, combined with the commonness of breast cancer, may

explain the recent unsuccessful attempts to identify ATM mutations inbreast cancer families. The situation is confounded by the difficulty ofdetecting all ATM mutations in a population, because the most efficient techniques currently available for the detection of A-T alíeles

require the mutant alíeleto be expressed as a truncated protein. Laakeand Borreson-Dale screened for the Norwegian founder mutationpresent in 55% of Norwegian A-T patients and found 1 A-T carrier in

447 consecutive breast cancer cases in Norway, as well as a carrier ofthe Norwegian founder mutation along with 2 additional A-T carriers

among 88 Swedish breast cancer cases (21). They are currentlylooking for LOH in these tumors, because the loss of the wild-type

alíeleshould be common if lack of ATM function is truly contributingto the development of cancer in these patients. Taking the complementary approach of assessing the status of relatives of A-T patients

where heterozygosity for a ATM mutation was confirmed, they founda relative risk for all cancers of 1.66 and a relative risk of 2.68 forbreast cancer in the Norwegan and Danish populations (22). In otherstudies of A-T relatives, Chessa found no evidence of increasedcancer risk, whereas Stoppa-Lyonnet supported an increased relativerisk of cancer in A-T carriers. Bendix presented preliminary data thatalso did not support a high frequency of A-T heterozygosity in

unselected breast cancer cases. These studies confirm that althoughthere is an increased risk of cancer in A-T carriers, this risk is small.

Wolman made the observation that ATM mutation seems to becorrelated with unusual morphology of breast tumors. Tumors in A-T

hétérozygotestended to be lobular and have colloid changes morefrequently than those of noncarriers, suggesting that it may be morefruitful to seek ATM mutations in specific types of breast tumors.Sturne looked at the level of ATM mRNA in breast cancer, benignbreast lesions, and normal breast tissue and found that normal tissueexpressed the highest levels oíATM, benign lesions expressed less,and breast carcinomas expressed the least ATM mRNA. Preliminarymutation analyses did not reveal any mutations in the ATM kinasedomain, but the results suggest that the ATM gene may be involved inneoplastic breast disease by an epigenetic mechanism.

Whereas ATM heterozygosity may not be a major genetic determinant in cancer, ATM seems to play a classic tumor suppressor role inT-prolymphocytic leukemia, which is highly associated with loss or

mutation of the ATM locus (23). This rare malignancy, although seenin young adult A-T patients (24), usually affects older patients and is

associated with somatic loss of both ATM alíelesin the tumor, ratherthan germ-line mutation of one alíeleand somatic mutation of the

second alíele.Similar results were presented by Schaffner, Stern,Yuille, and Vorechovsky. showing alíeleloss or mutation of ATM inthis rare cancer. Missense mutations were relatively more common inthese somatic events, rather than the predominance of truncatingmutations seen in A-T patients, although most of the mutations were

still predicted to be inactivating due to the change of conserved

3481

MEETING REPORT

residues in the active site of the ATM kinase (23). Vorechovsky erozygotes are more sensitive to DNA-damaging agents in vitro doessuggested that the incidence of ATM mutation in T-prolymphocytic

leukemia may be underestimated due to the inefficiency of mutationdetection for this large gene, and that ATM may be a key gene in themalignant transformation of T cells.

The idea that ATM might act as a tumor suppressor gene in othersporadic tumors is supported by the high frequency of LOH includingthe ATM locus at Ilq23.1 (25-27). Five groups presented data on

LOH at the ATM locus on 1Iq23.1 in tumors; these concerned breast,ovarian, and colon cancer. LOH at ATM affected 30-40% of breast

tumors. Rio showed data suggesting that ATM loss was associatedwith larger tumor size, whereas, in the same survey, LOH at BRCA1was associated more strongly with higher grade and negative hormonereceptors, suggesting that these genes act in different pathways intumorigenesis. In ovarian carcinoma, Launonen detected LOH at ATMin 61% of cases, with the majority simultaneously deleted at 1Iq23.2-

q25, where a second tumor suppressor locus is thought to be located.LOH at ATM was not correlated with prognosis, whereas the moredistal deletion was correlated with a more aggressive tumor. Granchosuggested that ATM loss is not common in colon tumors. The existence of a second tumor suppressor locus about 500 kb distal to ATMhas been suggested recently, based on LOH studies with closelyspaced markers (28). Given the large regions commonly deleted incases of LOH, it is therefore difficult to determine the relative importance of these two genes in any given tumor. It is essential tocontinue these studies by investigating the remaining alíelein cases ofLOH at ATM.

ATM and Radiation Sensitivity

It was first reported more than 30 years ago that A-T patients are

exquisitely sensitive to ionizing radiation. Tatsumi now reportsthat A-T cells exhibit a predominance of deletion mutations at a

test locus after irradiation. The abnormal response to irradiationseen in A-T homozygotes is often seen to a lesser degree in A-T

hétérozygotes,and it seems reasonable that hétérozygotesexpressing this defect may also be more prone to malignancy. It has longbeen one of the goals of A-T research to discover a way to detectA-T hétérozygotesin an effort to remove these sensitive individ

uals from the standard radiotherapy protocols and therefore deliverincreased and more effective doses to normal individuals withoutincreased morbidity. In keeping with this reasoning, radiationoverreactors might prove to be a cancer patient population that isrich in A-T hétérozygotes.Despite this, even though many A-Thétérozygotesare moderately radiosensitive in vitro, heterozygos-

ity at ATM does not account for a significant portion of radiotherapy overreactors, as presented by both Scott and Ramsay. Hallexamined the modulation of p53 protein levels in lymphoblastoidcell lines established from radiosensitive breast cancer patients andfound that the mean increase in p53 after exposure to ionizingradiation for the 16 cell lines examined to date was intermediatebetween that observed in A-T homozygote and normal cell lines,

indicating an alteration somewhere in the ATM signaling pathway.Tchirkov performed chromosome analysis on fresh lymphocytes

irradiated in the G2 phase with 1 Gy and found that 13 A-T hétérozy

gotes could be distinguished from normal controls by showing anincreased level of chromatid damage and suggested the use of this testto identify A-T hétérozygotesin cancer populations (29). Pernin alsofound that A-T hétérozygotescould be distinguished by their inter

mediate apoptotic response to etoposide and streptonigrin, again cautioning that A-T hétérozygotesmay have a tendency to overreact tochemotherapy. It is unclear why the consistent finding that the lym-phoblasts and fibroblasts of some or even the majority of A-T het-

not translate into finding an excess of A-T carriers in the group of

radiation and chemotherapy overreactors. The cellular responses toDNA damage are complex, and it cannot be expected that ATMmutation would be the only defect leading to radiation hypersensitiv-

ity; however, it seems that its effect in vivo is less than previouslypredicted.

A-T-related Disorders

Weemaes presented an overview of the NBS, which has beenconsidered a variant of A-T, primarily because some of the symptoms

of both diseases overlap, such as immunodeficiency, chromosomalinstability, hypersensitivity to ionizing radiation, and high cancer risk.NBS patients also manifest microcephaly and typical faciès,but donot have ataxia, telangiectasia, or elevated serum AFP. With thecloning of the ATM gene, the question of whether the NBS represented a true variant of A-T, possibly due to an unusual ATM muta

tion, could be addressed. It is now clear that NBS is caused by defectsin a gene distinct from ATM, and the genetic and functional mappingof the NBS gene to 8q21 was reported by Chrzanowska, Saar, Con-

cannon, and Komatsu (30, 31). While this report was in preparation,an NBS gene, NBS1, was cloned (32, 33), and the biochemicalrelationship between the two gene products can now be studied.

Clinical and biochemical studies of NBS have continued. Follow-

up of Polish NBS patients by Chrzanowska (34) revealed that ovarianfailure in NBS females is quite profound, but only a slight delay inpuberty was observed in males, and intelligence in preschool NBSchildren progressed from mild to moderate retardation by the age of14 years. Jongmans and Komatsu both investigated whether NBS cellsexhibited defects similar to those of A-T cells after irradiation (35,

36). p53 induction was attenuated after irradiation of NBS cells,although not as severely as that in A-T cells, as was transcriptionalactivation of WAF1/CIP1 (p21) and execution of the G,-S cell cycle

checkpoint. NBS cells also exhibited a prolonged accumulation ofcells in G2 after irradiation. A-T and NBS lymphoblastoid cell lines

exhibited similar levels of chromosomal damage in response to DNAstrand-breaking agents, but Stumm demonstrated that NBS cells dif

fered in their response to alkylating agents. These data, as well as theclinical findings, suggest that the NBS gene product performs similarfunctions to ATM, possibly acting in the same pathway.

In addition to NBS, several patients with some but not all of thehallmarks of A-T have been described and studied over the years,

typically presenting with a mild clinical course and/or intermediateradiosensitivity. van Beizen presented a family with three male siblings affected with a mild resting tremor, rearrangements of chromosomes 7 and 14 typical of A-T, and elevated AFP, but near normal

radiosensitivity and immunological tests. Linkage analysis of thisfamily was consistent with the ATM locus at Ilq23.1 being thedisease-causing gene in this family, although no mutations in the

coding or promoter regions of ATM have been found as yet.

ATM Functions

The phenotype of A-T cells in culture had been studied in detail for

many years before the cloning of the ATM gene. The speculation thatthe A-T gene product would be involved in the signaling pathway that

led to cell cycle checkpoints in the presence of DNA strand breaks andthat the pleiotropic phenotype would be due to defects in different endpoints in a complex web of signaling interactions is now beingconfirmed (37, 38). The cloning of the gene immediately led to abroader understanding of A-T functions, because ATM belongs to an

extended family of similar genes, and the phenotypes of defects inthese genes replicate subsets of the A-T phenotype in several model

3482

MEETING REPORT

systems, including fruit flies and yeast (5). An example of the functional overlap between these large protein kinases was given by Meyn,who showed that the yeast Tell protein could suppress spontaneoushyperrecombination and could partially suppress excessive apoptosisof A-T cells after irradiation.

Lavin and Kedar discussed their investigation of the cell cyclecheckpoint defects in A-T cells and the molecular pathways involved

(39). Their results and those of others suggest that in addition to p53,c-Abl is involved in executing the G,-S-phase checkpoint after irradiation, and that ATM binds constitutively to c-Abl and activates thekinase activity of c-Abl both in whole cells and in in vitro assays witha recombinant ATM phosphatidylinositol 3'-kinase domain fragment

(40, 41). Other interacting proteins are being identified through yeasttwo-hybrid screens or coimmunoprecipitation and suggest that the

molecules required for cell cycle progression are associated withseveral regulatory and inhibitor molecules in a large complex, and thatcell cycle progression may be regulated by cross-phosphorylation and

autophosphorylation of this complex.Uhrhammer and Chen showed that ATM fragments expressed in the

antisense orientation in normal cells could impose a radiosensitivephenotype (42).4 ATM antisense expression reduced ATM protein

levels in the normal cells, increased the number of chromosomeaberrations per metaphase, reduced cellular survival, and abrogatedcell cycle checkpoints after irradiation. This recreation of the A-T

phenotype will be useful for a wide variety of experiments in whichit is essential to have isogenic cell lines with and without ATMfunction. Uhrhammer showed that ATM antisense introduced into twoB-cell lymphoma cell lines could increase the radiosensitivity of the

cells regardless of their initial radiation survival and p53 status,suggesting that ATM antisense might be a useful gene therapy strategyfor cancer patients.

Humar et al. (43) extended the study of programmed cell death afterirradiation of A-T cells. A subset of A-T lymphoblastoid cells under

went apoptosis during the first 3 h after irradiation, although they didnot show any defect in double-strand break rejoining capacity. This

response was not dependent on activated p53, because activated p53could not be detected in the cells. This early apoptotic response is inaddition to the late apoptotic response of A-T lymphoblastoid cells,which occurs 2-5 days after irradiation, when the cells are arrested in

G2 (44). Although Humar et al. reported normal capacity for rejoiningDNA strand breaks, Tachibana reported reduced fidelity of repair,confirming the work of Thacker et al. (45).

Dantzer investigated the activity of PARP, an enzyme involved inthe detection of DNA strand breaks, the deficiency of which causesacute radiation sensitivity in mice (46). Given the similarities in thefunctions of ATM and PARP, it is possible that the two proteins actin concert or in the same pathway. A-T cells showed no deficiency in

PARP activity; however, immunoprecipitation showed an associationbetween ATM and PARP. Double knockout mice (PARP-/-;Atm—/—) are currently being made, and it will be interesting to see if

there is a synergistic effect of the two deficiencies.A key advance in the study of A-T, and one of great importance to

developing new therapeutic strategies for the patients, is the development of a mouse model by disruption of the murine Atm gene. AsWynshaw-Boris reviewed, the phenotype of Atm—/—mice is very

similar to that of human homozygotes: the mice are radiosensitive;they develop cancer at high frequency; they have immunodeficiency;and they are infertile (7, 8). There are also some differences in thephenotype of the Atm—/—mice; for example, although there is

histologically detectable degeneration in the brain, they do not seem

4 N. Zhang, P. Chen. K. K. Khanna. S. Scon, M. Gatei, S. Kozlov, D. Wallers, K.

Spring, T. Yen. and M. Lavin, unpublished observations.

to have ataxia or progressive cerebellar degeneration, only a subtledefect in coordination (47). Other differences include a more severemeiotic defect and the homogeneity of malignancy in the mice, with100% of Atm-/— mice developing T-cell lymphoma at an early age.

Ashley and Plug demonstrated the association of Atm and anothermember of the phosphatidylinositol 3'-kinase family, Atr, with the

synaptonemal complexes of chromosomes in meiosis I of the mouse(48). During early meiosis I, as the chromosomes begin to condense,Atr and Rad51 appear together with additional proteins in foci alongthe unsynapsed axial elements in both normal and Atm-/- mice.

These foci are known as early RNs and are thought to be involved inthe check for homology in preparation for reciprocal recombination.In normal mice, as soon as the chromosomes synapse, Atr leaves theRN and Atm appears in its place, along with the addition of RPA. Atmand RPA remain associated with the RN as it is processed into a lateRN, where DNA strand breaks and recombination are thought tooccur. Thus, it seems that Atm is present at the sites of DNA strandbreaks, presumably to monitor the correct processing of the break andensure an orderly progression through meiosis I. In Atm—/—mice,

pairing is initiated normally; however, without functional Atm at theRN, the synaptonemal complex begins to fragment, becoming progressively more abnormal, until the cell is morphologically apoptotic.Plug found that fragmentation often occurred at foci of RPA, indicating that it was the synapsed axes that were breaking, and implying thatthe DNA strand breaks that occur as an obligatory step of recombination were not held together by the maturing RN in the absence ofAtm.

McElligot showed that the signaling molecule Chkl/Rad27sp alsoassociated with meiotic chromosomes, although at a time later thanAtm and Atr, and that this did not occur in Atm-/- spermatocytes

(49). He also showed that the kinase activity of Atr was dependent onAtm. BRCA1 was found to colocalize with Atr on meiotic chromosomes, but its function was not dependent on Atm. Both McElligotand Meyn showed that apoptosis in Atm—/—spermatocytes was not

dependent on p53, in contrast to fibroblasts. Expression of p53 inspermatocytes of Atm-/- mice created in the Xu and Leder labs was

undetectable, whereas in normal spermatocytes, p53 was highly expressed. This result is in contrast to Atm-/- mice created by

Wynshaw-Boris, which may have a slightly more severe phenotype

and apparently express high levels of p53 in spermatocytes (50).These different results need to be confirmed and can perhaps beexplained by genetic background differences between the mousestrains or differences in the experimental protocols.

Diagnosis and Medical Aspects

A detailed immunohistochemical analysis of cerebellar changes inA-T patients was performed by Becker-Catania. She reported ectopiePurkinje cells in the granular cell and molecular layers of A-T patient

cerebellums as well as dysmorphic and degenerating axons similar tothose seen in elderly people and disorganization of parallel fibers inthe molecular layer. These findings suggested a developmental abnormality, with the loss of Purkinje cells being a later event. Thus, itseems that the primary lesion in the central nervous system of A-T

patients may not be in the cerebellum.Phenotypic heterogeneity in A-T was revisited by Sanai, who

documented differences in immunoglobulin deficiencies in a surveyof 56 patients and found minimal correlation between immunoglobulin patterns and susceptibility to infections. The immune defect inA-T was studied further by Zielen, who, like Sanai, reported an

impaired ability to generate IgG against polysaccharide vaccines inalmost all patients, whereas antibody responses to protein antigenswere normal. A-T lymphocytes showed increased levels of activated

3483

MEETING REPORT

CD4 cells and higher expression of the preapoptotic marker CD95(Fas/Apol), as shown by Schubert. This finding is consistent with themodel of ongoing cell loss in A-T. Regueiro and Porras also reported

elevated natural killer cell levels in a group of Spanish and CostaRican patients.

Throughout the meeting, the lack of obvious correlation betweenthe A-T molecular genotype and clinical phenotype (in particular, the

variation in immune defects, but also the variation in the course of thecerebellar degeneration) led the participants to discuss repeatedly howto define A-T. Aside from cases with a very clear A-T phenotype, it

seems difficult to decide whether using highly restrictive criteria ismore appropriate than using a broader clinical definition. It has beenshown that some variant patients have ATM mutations, but other rarepatients may have defects in other genes. This issue remains to beresolved.

In the end, the focus of the workshop shifted back to viewing thepatients as individuals needing medical care. The mission of theworkshop was not only to present the latest molecular findings andhypotheses, but to broaden awareness of the disease among cliniciansand discuss therapeutic strategies. Being a rare disorder, and one thatis not always easily diagnosed, family practitioners and pediatriciansare often not familiar with A-T. Many families could relate anecdotes

about misdiagnoses and inappropriate treatments that continued formonths, even years, before the correct diagnosis was made. Jabadodiscussed the management of A-T children. No cure or effective

treatment has yet been found, but it is generally agreed that individually tailored treatment, including aggressive physical therapy, antibiotics and supplemental immunoglobulins as necessary, and surveillance by a variety of specialists, was key to the quality of life for A-Tchildren. It seems that routine vaccination of A-T children is not

harmful and should be performed. Antioxidants are recommended.The progressive ataxia has the greatest impact on the patient and hisor her family, as the ability to communicate through speech, writing,drawing, and reading becomes severely impaired. To provide somemethod of predicting the neurological course of A-T patients, Craw

ford is developing a multidimensional clinical scale. It is as yetunclear what the nature of this neurological scale should be, but areliable and reproducible scale is required for assessing new therapies.

Future Directions

The field of A-T research has diverged explosively since the SixthInternational Workshop on A-T and the cloning of the ATM gene. In

1994, it was predicted that the Seventh International Workshop onA-T would focus squarely on gene therapy for A-T (51). Unfortu

nately, ATM gene therapy is impractical at this time, but the unraveling of the role of ATM in DNA damage signaling opens other doorsfor research and therapy. In the future, it is expected that a betterunderstanding of the variability of the A-T phenotype will be reached,

possibly using a refined definition of the disease. We also expect tobetter define the epidemiological and molecular roles of ATM insporadic and familial cancer and perhaps to use ATM antisense as atherapeutic tool to combat various cancers. It is hoped that the EighthInternational Workshop on A-T will provide some of these answers.

Acknowledgments

The attendees thank the members of the Laboratoire d'Oncologie Molécu

laire for their hospitality.

References1. Gatti, R. A. A-T. In: B. Vogelstein and K. W. Kinzler (eds.). The Genetic Basis of

Human Cancer, pp. 275-300. New York: McGraw-Hill, 1998.2. Sedgwick, R. P., and Boder, E. Ataxia-telangiectasia. In: P. }. Vinken, G. W. Bruyn.

and H. L. Klawans (eds.). Handbook of Clinical Neurology, pp. 347-423. Amsterdam: Elsevier Scientific Publishers, 1991.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Jaspers. N. G. J.. Gatti. R. A.. Baan. C.. Linssen. P. C. M. L.. and Bootsma. D. Geneticcomplementation of ataxia-telangieclasia and Nijmegen breakage syndrome: a surveyof 50 patients. Cytogenet. Cell Genet., 49: 259-263. 1988.Gatti, R. A., Berkel, I.. Boder, E.. Braedt, G., Charmley. P.. Concannon, P., Ersoy, F.,Foroud, T., Jasper, N. G. J., Lange. K.. Lahtrop. G. M., Leppert. M.. Nakamura. Y..O'Connell. P.. Paterson. M., Salser, W., Sana!. O., Silver, J., Sparkes. R. S., Susi. E.,

Weeks. D. E.. Wei. S.. White. R.. and Yoder, F. Localization of an ataxia-telangiec-tasia gene to chromosome llq22-23. Nature (Lond.), 336: 577-580, 1988.Savitsky. K., Bar-Shira. A.. Gilad. S.. Rotman. G., Ziv, Y.. Vanagaite, L.. Tagle,

D. A., Smith, S., Uziel, T., Seez, S., Ashkenazi, M., Pecker, I., Frydman. M.. Hamik.R., Patanjali, S. R., Simmons, A., Clines, G. A.. Sanici. A., Gatti. R. A., Chessa, L.,Sanai, O., Lavin, M. F., Jaspers. N. G. J.. Taylor, A. M. R., Arieti, C. F.. Miki. T..Weissman, S. M., Lovett, M.. Collins, F. S., and Shiloh. Y. A single alaxia-

lelangieclasia gene with a product similar to PI3 kinase. Science (Washington DC),268: 1749-1753. 1995.

Zhang. N.. Chen, P.. Khanna, K. K.. Scott. S.. Gatei. M.. Kozlov, S.. Walters, D.,Spring, K., Yen, T., and Lavin, M. Proc. Nati. Acad. Sci. USA. 94: 8021-8026, 1997.Barlow. C., Hirolsune. S.. Paylor. R.. Liyanage, M.. Eckhaus. M.. Collins, F.. Shiloh.Y.. Crawley, J. N., Ried, T., Tagle. D.. and Wynshaw-Boris, A. ATM-deficient mice,a paradigm of alaxia-lelangiectasia. Cell, 86: 159-171, 1996.

Xu, Y., and Baltimore. D. Dual roles of ATM in the cellular response to radiation andin cell growth conlrol. Genes Dev.. 10: 2401-2410. 1996.Ziv, Y., Bar-Shira, A.. Pecker, I., Russell. P.. Jorgensen. T. J., Tsarfaty, I., and Shiloh.Y. Recombinant ATM protein complements the cellular A-T phenolype. Oncogene./5: 159-167, 1997.Kaslan, M. B.. Zhan, Q., and El-Deiry, W. A mammalian cell cycle checkpointpalhway utilizing p53 and GADD45 is defective in ataxia-telangiectasia. Cell. 71:587-597, 1992.Khanna. K. K.. Beamish. H.. Yan, J.. Hobson. K.. Williams, R., Dunn, I., and Lavin,M. F. Nature of G,/S cell cycle checkpoint defect in ataxia-telangiectasia. Oncogene.//: 609-618, 1995.McConville, C. M.. Stankovic. T., Byrd. P. J.. McGuire, G., Yao, Q-Y., Lennox,

G. G., and Taylor, A. M. R. Mutations associated with variant phenotypes inataxia-telangiectasia. Am. J. Hum. Genel., 59: 320-330. 1996.

Gilad. S.. Chessa, L.. Khosravi, R., Russell. P., Galanty, Y.. Piane. M., Gatti, R. A.,Jorgensen. T. J.. Shiloh. Y., and Bar-Shira. A. Genotype-phenotype relationships inataxia-telangiectasia (A-T) and A-T variants. Am. J. Hum. Genet., 62: 551-561.

1998.Teletar. M.. Teraoka, S.. Wang, Z.. Chun, H.. Liang. T., Castellvi-Bel, S.. Udar, N..Borresen-Dale, A-L., Chessa, L., Bernatowska-Matuzkiewies. E., Porras, O.,Watanabe. M., Junker, A.. Concannon, P., and Gatti, R. A. Alaxia-lelangiectasia:

identificalion and deleclion of founder effecl mulalions in Ihe ATM gene in elhnicpopulations. Am. J. Hum. Genet., 62: 86-97, 1998.Shiloh, Y., and Rotman. G. Ataxia-telangieclasia and Ihe ATM gene: linking neuro-degeneralion, immunodeficiency and cancer to cell cycle checkpoint. J. Clin. Immu-nol.. 16: 254-260. 1996.

Toyoshima, M., Hará,T., Zhang, H., Yamamoto. T., Akaboshi. S.. Nanba. E., Ohno.K., Hori, N.. Salo, K.. and Takeshila. K. Ataxia-lelangieciasia wilhoul immunodeficiency: novel point mulalions wilhin and adjaccnl to the phosphatidylinosilol 3-kinase-like domain. Am. J. Hum. Genel., 75: 141-144. 1998.

Swift. M., Morrell, D., Massey, R. B., and Chase, C. L. Incidence of cancer in 161families affecled by alaxia-lelangieclasia. N. Engl. J. Med., 325: 1831-1836, 1991.Easlon. D. F. Cancer risks in alaxia-lelangieclasia helerozygoles. Ini. J. Radial. Biol..66: S177-S182. 1994.

FilzGerald, M., Bean, J. M.. Hegde, S. D., Unsal, H., MacDonald, D.. Harkin, D. P.,Finkelslein, D. M., Isselbacher, K. J.. and Haber. D. A. Heterozygous ATM mulalionsdo noi contribute lo early onsel of breasl cancer. Nal. Genel., 15: 307-310. 1997.Alhma, P.. Rappapon. R.. and Swift. M. Molecular genolyping shows lhal alaxia-

lelangieclasia helerozygotes are predisposed lo breasl cancer. Cancer Genel.Cylogenet., 92: 130-134, 1996.

Vorechovsky, I., Luo, L., Lindblom, A., Negrini, M., Webster, A. D. B., Croce, C. M.,and Hammarslrom. L. ATM mulalions in cancer families. Cancer Res., 56: 4130-

4133, 1996.Borreson, A. L., Anderson, T. I., Trelli, S., Heiberg, A., and Moller, P. Breasl cancerand olher cancers in Norwegian families wilh alaxia-lelangieclasia. Genes Chromosomes Cancer, 2: 339-340, 1990.

Vorechovsky, I., Luo, L., Dyer, M. J. S., Calovsky, D.. Amlol, P. L., Yaxley, J. C.,Foroni. L.. Hammarslrom. L.. Webster. A. D., and Yuille. M. A. Clustering ofmissense mulalions in Ihe alaxia-lelangieclasia gene in sporadic T-cell leukemia. Nat.Genel., 17: 96-99, 1997.

Taylor. A. M. R., Melcalfe, J. A., Thick, J., and Mak, Y. F. Leukemia and lymphomain alaxia-lelangieclasia. Blood, X7: 423-438, 1996.

Cabra, H., Walson, J. E. V.. Taylor, K. J., Mackay. J.. Leonard, R. C. F., Steel, C. M.,Porleous, D. J.. and Smylh. J. F. Definilion and refmemenl of a region of loss ofhelerozygosity al 1Iq23.3-q24.3 in epilhelial ovarian cancer associated wilh poorprognosis. Cancer Res., 56: 950-954, 1996.

Winqvist, R., Hamplon, G. M., Mannermaa, A., Blanco, G., Alavaikko. M..Kiviniemi, H., Taskinen, P. J., Evans, G. A., Wrighl, F. A., Newsham, 1., andCavanee, W. K. Loss of helerozygosily for chromosome 11 in primary humanbreasl lumors is associated wilh poor survival after metastasis. Cancer Res., 55:2660-2664. 1995.Hamplon, G. M., Mannermaa. A., Winqvisl, R., Alavaikko, M., Blanco, G., Taskinen.P. J., Kiviniemi, H., Newsham. I.. Cavanee, W. K.. and Evans. G. A. Loss ofhelerozygosily in sporadic human breasl carcinoma: a common region belween 1Iq22and 1Iq23.3. Cancer Res., 54: 4586-4589, 1994.

3484

MEETING REPORT

28. Negrini. M.. Rasio, D.. Hampton, O. M., Sabbioni, S., Rallan, S.. Carter, S. L.,Rosenberg, A. L., Schwartz, G. F., Shiloh, Y., and Cavanee. W. K. Definition andrefinement of chromosome 11 regions of loss of heterozygosity in breast cancer:identification of a new region at Ilq23.3. Cancer Res., 55: 3003-3007, 1995.

29. Tchirkov, A., Bay, J-O., Pernin, D., Bignon, Y-J., Rio, P.. Granchio, M.. Kwiatkowski,F.. Giollant, M., Malet, P.. and Verrelle. P. Detection of heterozygous carriers of theataxia-telangiectasia (ATM) gene by G2 phase chromosomal radiosensitivity ofperipheral blood lymphocytes. Hum. Genet., 101: 312-316, 1997.

30. Saar, K., Chrzanowska, K. H., Stumm, M., Jung, M.. Nürnberg,G.. Wienker. T. F.,Seemanova, E., Wegner, R. D., Reis, A., and Sperling. K. The gene for the ataxia-telangiectasia variant, Nijmegen breakage syndrome, maps to a 1 cM interval onchromosome 8q21. Am. J. Hum. Genet., 60: 605-610, 1997.

31. Matsuura, S., Weemaes. C.. Smeets. D., Takami, H., Kondo, N., Sakamoto. S.,Yano, N., Tauchi, H., Endo. S.. Oshimura. M., and Komatsu, K. Genetic mappingusing microcell mediated chromosome transfer suggests a locus for Nijmegenbreakage syndrome al chromosome 8q21-24. Am. J. Hum. Genet., 60: 1487-1497, 1997.

32. Varón,R., Vissinga. C., Platzer, M., Cerosaletti. K. M., Chrzanowska, K. H., Saar, K.,Beckmann, G., Seemanova, E., Cooper, P. R., Nowak, N. J., Stumm, M., Weemaes,C. M. R., Gatti, R. A., Wilson, R. K.. Digweed, M.. Roscnthatl. A.. Sperling. K..Concannon, P.. and Reis, A. Nibrin. a novel DNA double-strand break repair prolein,is mutated in Nijmegen breakage syndrome. Cell. 93: 467-476. 1998.

33. Carney, J. P.. Maser, R. S., Olivares, H., Davis, E. M., Le Beau, M., Yales, J. R.,Hayes, L., Morgan, W. F.. and Petrini, J. H. J. The hMrel l/hRad50 protein complexand Nijmegen breakage syndrome: linkage of double-strand break repair to thecellular DNA damage response. Cell, 93: 477-486, 1998.

34. van der Bürgt,I., Chrzanowska, K. H., Smeets, D., and Weemaes. C. Nijmegenbreakage syndrome. J. Med. Genet., 33: 153-156, 1996.

35. Jongmans, W., Vuillaume, M., Chrzanowska. K., Smeets, D., Sperling, K.. and Hall.J. Nijmegen breakage syndrome cells fail lo induce the p53-mediated DNA damageresponse following exposure to ionizing radiation. Mol. Cell. Biol.. 17: 5016-5022,

1997.36. Matsuura. K., Balmukhanor. T.. Tauchi, H., Weemaes, C.. Smeets, D.. Chrzanowska.

K. H., Endou, S., Malsuura. S., and Komatsu, K. Radiation induction of p53 in cellsfrom Nijmegen breakage syndrome is defective but not similar to ataxia-telangiectasia. Biochem. Biophys. Res. Commun., 242: 602-607, 1998.

37. Meyn. M. S. Ataxia-telangiectasia and cellular response to DNA damage. CancerRes.. 55: 5991-6000. 1995.

38. Shiloh, Y. Ataxia-telangiectasia: closer to unraveling the mystery. Eur. J. Hum.Genet., 3: 116-138, 1995.

39. Wallers, D.. Khanna, K. K., Beamish, H.. Birrell, G., Spring, K., Kedar. P., Gatei. M.,Stenzel, D., Hobson, K., Kozlov, S., Zhang. N.. Farrell. A.. Ramsay. J., Gatti, R.. and

Lavin, M. Cellular localization of the ataxia-telangiectasia gene product and discrimination between mulaled and normal forms. Oncogene, 14: 1911-1921. 1997.

40. Shafman. T.. Khanna, K. K., Kedar, P., Spring. K.. Kozlov. S.. Yen, T., Hobson, K.,Galei. M., Zhan, W.. Walters. D.. Egerton, M., Shiloh, Y.. Kharbanda, S., Kufe. D.,and Lavin. M. F. Interaction between ATM protein and c-Abl in response to DNAdamage. Nature (Lond.), 387: 520-523, 1997.

41. Baskaran, R., Wood, L. D.. Whitaker, L. L.. Canman, C. E., Morgan, S. E., Xu, Y.,Barlow, C., Baltimore, D., Wynshaw-Boris, A., Kastan, M. B., and Wang. J. Y.Ataxia-telangiectasia mutant protein activates c-Abl tyrosine kinase in response toionizing radiation. Nature (Lond.), 387: 516-519, 1997.

42. Uhrhammer. N.. Frilz. E., Boyden. L.. and Meyn, M.S. ATM anlisense veclor confersan alaxia-telangiectasia phenolype on normal cells. Am. J. Hum. Genet., 59: 10, 1996.

43. Humar, B.. Muller, H., and Scoli, R. J. Elevated frequency of p53-independentapoptosis after irradiation increases levels of DNA breaks in ataxia-telangiectasialymphoblasts. Int. J. Radial. Biol., 72: 257-269, 1997.

44. Meyn, M. S., Slrasfeld, L., and Allen. C. Testing the role of p53 in genetic instabilityand apoptosis in alaxia-telangiectasia. Int. J. Radial. Biol.. 66: 141-149. 1994.

45. Thacker, J., Chalk, J.. Ganesh, A., and Thacker. J. A mechanism for deletionformation in DNA by human cell extracts: the involvemenl of short sequence repeats.Nucleic Acids Res.. 20: 6183-6188, 1992.

46. de Muclia, J. M., Niedergang, C., Trucco, C., Ricoul, M., Dulrillaux, B., Mark, M..Oliver, F. J., Masson, M.. Dierich, A.. Leleur, M., Walztinger. C., Chambón. P.. andde Murcia, G. Requirement of poly(ADP-ribose) polymerase in recovery from DNAdamage in mice and cells. Proc. Nati. Acad. Sci. USA, 94: 7303-7307, 1997.

47. Kuljis, R. O., Xu, Y., Õguila. M. C., and Baltimore, D. Degeneration of neurons,synapses, and neuropil and glial activation in a murine Atm knockoul model ofataxia-lelangieclasia. Proc. Nail. Acad. Sci. USA, 94: 12688-12693, 1997.

48. Keegan, K. S., Holtzman, D. A., Plug, A. W., Chrislenson, E. R., Brainerd, E. E.,Flaggs, G., Bentley. N. J., Taylor, E. M., Meyn, M. S., Moss, S. B., Carr. A. M.,Ashley, T., and Hoekstra, M. F. The Atr and Atm protein kinases associale withdifferent sites along meiolically pairing chromosomes. Genes Dev., 10: 2423-2437,

1996.49. Flaggs, G., Plug, A. W.. Dunks, K. M., Mundl. K. E., Ford, J. C., Quiggle, M. R..

Taylor. E. M., Westphal, C. H., Ashley, T., Hoekstra, M. F., and Carr, A. M.Aim-dependent interactions of a mammalian chkl homolog with meiolic chromosomes. Curr. Biol., 7: 977-986, 1997.

50. Barlow, C., Brown, K. D., Deng. C-X., Tagle. D. A., and Wynshaw-Boris, A. Atmselectively regulates distinct p53-dependenl cell-cycle checkpoini and apopiotic pathways. Nat. Genet., 17: 453-461, 1997.

51. Gatti, R. A., McConville, C. M., and Taylor, A. M. R. Sixlh International Workshopon Ataxia-Telangiectasia. Cancer Res., 54: 6007-6010. 1994.

3485