Identification of an alternative 5'-untranslated exon and new polymorphisms of angiotensin-converting enzyme 2 gene: lack of association with SARS in the Vietnamese population.

We analyzed genetic variations of angiotensin-converting enzyme 2 (ACE2), considering that it might influence patients' susceptibility to severe acute respiratory syndrome-associated coronavirus (SARS-CoV) or development of SARS as a functional receptor. By cloning of the full-length cDNA of the ACE2 gene in the lung, where replication occurs on SARS-CoV, it was shown that there are different splicing sites. All exons including the new alternative exon, exon-intron boundaries, and the corresponding 5'-flanking region of the gene were investigated and 19 single nucleotide polymorphisms (SNPs) were found. Out of these, 13 SNPs including one non-synonymous substitution and three 3'-UTR polymorphisms were newly identified. A case control study involving 44 SARS cases, 16 anti-SARS-CoV antibody-positive contacts, 87 antibody-negative contacts, and 50 non-contacts in Vietnam, failed to obtain any evidence that the ACE2 gene polymorphisms are involved in the disease process in the population. Nevertheless, identification of new 5'-untranslated exon and new SNPs is considered helpful in investigating regulation of ACE2 gene expression in the future. Copyright (c) 2005 Wiley-Liss, Inc.

INTRODUCTION

Severe acute respiratory syndrome (SARS) is an emerging infectious disease characterized by systemic inflammation followed by atypical pneumonia [Peiris et al., 2003b]. Shortly after the initial worldwide outbreak in 2003, SARS‐associated coronavirus (SARS‐CoV) was discovered as an etiological agent of SARS [Drosten et al., 2003; Ksiazek et al., 2003; Kuiken et al., 2003; Peiris et al., 2003a], and then angiotensin‐converting enzyme 2 (ACE2) was identified as a functional receptor of this newly arrived virus [Li et al., 2003]. More recently, CD209L was reported as being another alternative receptor for the virus, but it appears to be a less efficient entry site than ACE2 [Jeffers et al., 2004].

Virus receptors generally play a key role in the entry of the pathogen into the host cells and may influence development or progression of viral diseases. For example, it is well known that genetic polymorphism of chemokine receptor 5 (CCR5), a co‐receptor for human immunodeficiency virus‐1 (HIV‐1), influences the natural history of HIV‐1 infection. The mutant allele CCR5‐Δ32 does not produce a functional protein and has been shown to protect host cells against HIV‐1 infection, and progression into acquired immunodeficiency syndrome is delayed after seroconversion takes place [Dean et al., 1996; Liu et al., 1996; Samson et al., 1996]. By analogy with the above, we considered that genetic polymorphisms of ACE2 could influence SARS‐CoV infection or clinical manifestations of SARS.

ACE2 is a homologue of ACE1 and exhibits 40% identity of amino acid sequence to its N‐ and C‐terminal domains [Tipnis et al., 2000]. Similar to ACE1, ACE2 is a metalloprotease that constitutes a renin‐angiotensin system. Human full‐length ACE2 cDNAs have been cloned already from lymphoma (GenBank accession No. AF241254) [Tipnis et al., 2000], cardiac left ventricle (AF291820) [Donoghue et al., 2000] and testis (AY623811) [Douglas et al., 2004]. Based on published data, it has been said that the ACE2 gene (ACE2) contains 18 exons, and spans approximately 40 kb of genomic DNA on the human X‐chromosome. Although ACE2 mRNA expressions were demonstrated in the lung by the method of quantitative reverse transcription‐PCR (RT/PCR) [Harmer et al., 2002] and its protein expression was obviously shown by immunohistochemistry [Hamming et al., 2004], full‐length ACE2 cDNA has not been cloned from the lung so far. This is considered to be very likely as being an important replication site of SARS‐CoV [Haagmans et al., 2004].

In the present study, we attempted a full‐length cloning of ACE2 cDNA from the human lung and found a new alternative, the 5′‐untranslated exon. During this process, an extended region of the original exon 1 was identified in the testis' RNAs. Then, we explored genetic polymorphisms within 19 exons including new regions and the 5′‐flanking region of ACE2 and tried to determine whether the polymorphisms of ACE2 are associated with SARS in Vietnamese.

MATERIALS AND METHODS

Cloning of ACE2 cDNA From the Lung

Cloning was performed by combination of RT/PCR and 5′‐ and 3′‐ rapid amplification of cDNA ends (RACE) procedures, using human lung total RNA (Stratagene, La Jolla, CA) and human testis total RNA (Stratagene) as a control. The total RNAs were reverse‐transcribed using SuperScript III Reverse Transcriptase (Invitrogen, Carlsbad, CA) with oligo(dT)12–18, and then cDNA was amplified using Platinum Taq DNA Polymerase High Fidelity (Invitrogen) with primers ACE2‐exon 1s (5′‐CAA AGG CTG ATA AGA GAG AA‐3′) and ACE2‐exon 18 as (5′‐GAA CAG AAG TCA AAT CCA GA‐3′) to amplify the transcript of 2721 bp encompassing the original 18 exons of ACE2 gene on database.

The First Choice RLM‐RACE Kit (Ambion, Austin, TX) was used for 5′‐ and 3′‐RACE procedures following the manufacturer's recommendation. Gene‐specific primer sets for 5′‐RACE were ACE2‐5′Outer1 and ACE2‐5′Inner1 (5′‐GTG GAT ACA TTT GGG CAA GT‐3′ and 5′‐CCT AGA CTA AAA CCT CCT CA‐3′), and ACE2‐5′Outer2 and ACE2‐5′Inner2 (5′‐GAA GTA AGA AAG CCT CCA CA‐3′ and 5′‐CTC CTG ATC CTC TGT AGC CA‐3′). Gene specific primer set for 3′‐RACE was ACE2‐3′Outer and ACE2‐3′Inner (5′‐CAA TGA TGC TTT CCG TCT GA‐3′ and 5′‐ACA CTT GGA CCT CCT AAC CA‐3′). Nucleotide sequences of PCR products were directly determined by the automated DNA sequencer (PRISM 3100 Genetic Analyzer, Applied Biosystems, Foster City, CA).

To investigate expression of the exons on the 5′ side, RT/PCR procedures were performed on the total RNAs of human lung, testis, trachea (Stratagene), primary‐cultured bronchial epithelial cells [Lechner and LaVeck, 1985], small intestine (Ambion), and on the human major organ cDNAs (Bio Chain Institute) with the sense primer New‐exon (5′‐TTC TTA CTT CCA CGT GAC CT‐3′) or Extended‐exon 1 (5′‐GCT CAG CAG ATT GTT TAC TG‐3′) and the antisense primer ACE2‐5′Outer1.

Genomic DNA Samples for the Association Study

An association study between SARS patients and controls was reviewed and approved by local ethics committees. Of 62 cases fulfilling the World Health Organization case definition of probable SARS in Vietnam [WHO, 2003], 5 fatal cases and 3 non‐Vietnamese cases were excluded from this study. In the remaining 54 cases, 44 individuals agreed to participate in this study as cases. One hundred and three Vietnamese staff members, who did not develop SARS but may have come in contact with SARS patients in the hospital where nosocomial infection of SARS had arisen, were enrolled as contacts. Furthermore, 50 medical staff members who had been working in a separate building and those considered having no history of contact with SARS patients joined in this study as non‐contacts, according to information obtained by questionnaire. Peripheral blood samples of all the subjects were collected and genomic DNA was extracted from the blood cells by a method described elsewhere [Wang et al., 1994].

Testing for Antibody Response to the SARS‐CoV

To detect the antibody to the SARS‐CoV in serum, all the blood samples were tested with SARS ELISA (Genelabs Diagnostics Pte. Ltd., Singapore Science Park, Singapore) in accordance with the manufacturer's recommendation [Guan et al., 2004].

Identification of Polymorphisms Within ACE2 Gene

Of the 44 SARS cases and 103 contacts recruited, a half of the samples were randomly selected for searching polymorphisms within the ACE2 gene. PCR primers were designed to amplify 19 exons including the new alternative exon, exon‐intron boundaries and approximately 1,000 bp of the 5′‐flanking region of the new exon, reaching 2,000 bp upstream of the 5′‐end of the original exon 1 (Table I). Genomic DNA of each sample was subjected to PCR amplification followed by direct sequencing.

Table I

Primers Used to Identify Polymorphisms Within the ACE2 Gene

Region	Primer name	Primer sequence (5′–3′)	Product size
5′ flanking region	ACE2‐pro‐1‐sense	TAA TTC AGT CAG TGC TTG C	676 bp
	ACE2‐pro‐1‐anti	AAT AGT GGA GGC ATA GAT AAA
5′ flanking region	ACE2‐pro‐2‐sense	TTT GTG AGC TGC TTT ATT TT	618 bp
	ACE2‐pro‐2‐anti	TGC CAG AGT GTA TGT ATG AG
New alternate exon	ACE2‐new‐sense	TTA TTG CAA TGT CAC CTG A	470 bp
	ACE2‐new‐anti	TTA TGA CTA CTC TCC ACT CCA
5′ flanking region	ACE2‐pro‐3‐sense	TTT GAA TAG GTA AGT GAA GG	669 bp
	ACE2‐pro‐3‐anti	TAG AAC TAG GGA TCA TGA AGA
5′ flanking region	ACE2‐pro‐4‐sense	TGA ATT CCA TAA AGA CAA GG	653 bp
	ACE2‐pro‐4‐anti	AAA CTT GTC CAA AAA TGT CTT
Exon 1	ACE2‐ex1‐sense	ATC TTT AAC AGC TTT CTA GGA	644 bp
	ACE2‐ex1‐anti	AAC ATC CAA TCT CAC AAC TC
Exon 2	ACE2‐ex2‐sense	AAC TCA TCT ATG TCA CAG CAC	636 bp
	ACE2‐ex2‐anti	AAA TTA TAT GGA CAC CTT ACC
Exon 3	ACE2‐ex3‐sense	ACT TCT TTG GGT TTT GGT AG ACA TCA GGT CAT AAA GTG	627 bp
	ACE2‐ex3‐anti	GTT
Exon 4	ACE2‐ex4‐sense	TCA TTT CAG TGG TTT ATT TTC	521 bp
	ACE2‐ex4‐anti	CTT TTC TTT TTC CCC AGT A
Exon 5	ACE2‐ex5‐sense	CTT GTA TGG TTC TTG TGC TT	535 bp
	ACE2‐ex5‐anti	GGG CTG TCC TAT TAT TCT CTA
Exon 6	ACE2‐ex6‐sense	ACC TGT GTT CTC CCA AGT A	568 bp
	ACE2‐ex6‐anti	CTT TAT CAT TTG AAT TGC AG
Exon 7	ACE2‐ex7‐sense	TCA CCA AGT TAA GTA CAC GAA	562 bp
	ACE2‐ex7‐anti	TAC ACC TGC AAT TCA AGT TAT
Exon 8	ACE2‐ex8‐1‐sense	TTG CAG TGA GAA CAT TTG AAA	560 bp
	ACE2‐ex8‐1‐anti	CCT CTG TTG TCT CCC ATT T
Exon 8	ACE2‐ex8‐2‐sense	GCT GTG CAG TAG ATC TCA AA	643 bp
	ACE2‐ex8‐2‐anti	CAG ATT GTC CAC AGG TTC A
Exon 9	ACE2‐ex9‐sense	CTA TGA GCA AGA GAA CAG G	577 bp
	ACE2‐ex9‐anti	TCA CCA GTA GTA ATT TCC AGT
Exon 10	ACE2‐ex10‐sense	AGG GAG GAA ACT GAA ACT AAT	587 bp
	ACE2‐ex10‐anti	GGT ATC CAA ATG GAG ACT AAA
Exon 11	ACE2‐ex11‐sense	GTG CAC ACC TAT AAA CCA AG	615 bp
	ACE2‐ex11‐anti	TGA GCA TGT TTA GGG TAG AC
Exon 12	ACE2‐ex12‐sense	GTG AAA GGG CTA TTA ATC TGT	612 bp
	ACE2‐ex12‐anti	GAG AGG GCT GTA GTT ATG A
Exon 13	ACE2‐ex13‐sense	CAG GAA CCT AGA CCA TAC AA	636 bp
	ACE2‐ex13‐anti	GTT GCT TTC ACT ATG TCT CA
Exon 14	ACE2‐ex14‐sense	GTA CAA ATT AGG TCA TGG C	550 bp
	ACE2‐ex14‐anti	GAC GAG AGT CAA TTG AAA G
Exon 15	ACE2‐ex15‐sense	ATT ATT GGG TTT CAT CTC G	637 bp
	ACE2‐ex15‐anti	TAT AGG TCA ATG AAG GCA G
Exon 16	ACE2‐ex16‐sense	CAG AAC AAA TAG TGC CAA A	610 bp
	ACE2‐ex16‐anti	CAT AGT GGT AAC TTG CTT GAT
Exon 17	ACE2‐ex17‐sense	GCT CTG TCA CCT AGG CTA G	633 bp
	ACE2‐ex17‐anti	CTA GGA AGA TGA ACT GCT GAT
Exon 18	ACE2‐ex18‐1‐sense	TTA AGA TGA ATC CTA GCA GTG	655 bp
	ACE2‐ex18‐1‐anti	CAT TTA GAT TAT CCC TGA ACA
Exon 18	ACE2‐ex18‐2‐sense	TCT GGA TTT GAC TTC TGT TC	623 bp
	ACE2‐ex18‐2‐anti	AAC ACT GTG AGC AAA TAC AAA
Exon 18	ACE2‐ex18‐3‐sense	GAA CAG GTA GAG GAC ATT G	531 bp
	ACE2‐ex18‐3‐anti	GGG TAG TGA CTG TGA GAA ATA

Genotyping of Identified Polymorphisms

Non‐synonymous nucleotide substitutions and other variations with a minor allele frequency higher than 0.05 were subjected to genotyping in all SARS cases, contacts and non‐contacts. Consequently, one novel non‐synonymous substitution, two possible non‐synonymous polymorphisms in the database (dbSNP identification nos. rs4646116 and rs11798104), and variations of 3′‐UTR in exon 18 (position 39844) and of intron 3 (rs2285666, position 8789) were genotyped by the combination of direct sequencing method and single‐strand conformation polymorphism (SSCP) analysis or PCR‐based restriction fragment length polymorphism (RFLP) analysis.

Statistical Analysis

Disease associations were assessed by the chi‐square test. The P values less than 0.05 were considered significant in all the tests and data analysis was carried out using JMP version 5 (SAS Institute, Inc., Cary, NC).

RESULTS

Full‐Length ACE2 cDNAs From the Lung and Expression of the Transcripts

By the use of the RT/PCR encompassing all known exons of ACE2 and 3′‐RACE method, we could amplify ACE2 cDNA as PCR fragments completely corresponding to the published sequence of ACE2 cDNA (AF241254). The 5′‐RACE procedure on the total RNA of the lung demonstrated the presence of a new alternative exon (registered as AB193259), which consisted of a segment between position −1141 and −942 and was connected to the 5′‐end of the original exon 1. The 5′‐end of transcripts was extended to position −1141 repeatedly by both sets of gene‐specific primers. In addition, novel 65 nucleotides on the 5′‐side (registered as AB193260), extending the 5′‐end of the original exon 1 upstream, were amplified from the total RNA of testis. A schematic diagram of the exon‐intron structure is shown in Figure 1.

Figure 1

A schematic diagram of the ACE2 gene structure and the positions of SNPs. The known exons are depicted as open boxes. A solid box and a striped box indicate the new exon and the new extended region of the exon 1, respectively. The arrows represent locations of the SNPs analyzed in a case‐control study. The broken line depicts an alternative‐splicing site.

RT‐PCR revealed that the expression of the new alternative exon could be seen not only in the lung but also in the testis, trachea, bronchial epithelial cells, small intestine, and various major organs (data not shown). The new extended region was expressed not only in the testis but also in other organs including bronchial epithelial cells and the small intestine (data not shown).

Subgrouping of Subjects Based on the Status of Anti‐SARS‐CoV Antibody

Basic characteristics and sub‐grouping of subjects are shown in Table II. The 44 SARS cases, 103 contacts, and 50 non‐contacts were analyzed in the present study. Based on anti‐SARS‐CoV antibody titer in serum, the contacts were further divided into two subgroups, antibody‐positive contacts, and antibody‐negative contacts (data not shown).

Table II

Demographic Findings of Subjects and Subgroups

Groups	SARS cases (n = 44)	Contacts (n = 103)	Anti‐SARS‐CoV antibody		Non‐contacts (n = 50)
			Positive (n = 16)	Negative (n = 87)
Age (years), mean [range]	39.3 [17–76]	36.5 [15–68]	36.0 [25–50]	36.6 [15–68]	—a
Male/female	13/31	46/57	7/9	39/48	17/33

Data not available.

All exons including the new exon, exon‐intron boundaries and the corresponding 5′‐flanking region of ACE2 were tested to identify variations of ACE2 among SARS cases and contacts. As shown in Table III, 19 single nucleotide polymorphisms (SNPs) were identified. Six of them have already registered on dbSNP database, and 13 SNPs including one non‐synonymous substitution, from asparagine to serine at 638 (N638S) in the exon 15 (position 33205) were identified. All SNPs but one in intron 3 (rs2285666, position 8789) and another in exon 18 (position 39844) were found to be considerably rare among both SARS cases and contacts tested. In subsequent analysis, we therefore chose polymorphisms, and analyzed possible non‐synonymous substitution, excluding rare non‐coding variants among SARS patients and contacts.

Table III

SNPs Within the ACE2 Gene

Region	Positiona	dbSNP rs# cluster ID	Change of nucleotide (major/minor allele)	Change of amino acid (major/minor allele)	No. of individuals who had the minor allele
					SARS cases	Contacts
5′ flanking region	−751	NEWb	C/T	—	1	1
5′ flanking region	−671	NEW	G/A	—	1	1
5′ flanking region	−634	NEW	C/G	—	1	0
Intron 3	8789	rs2285666	A/G	—	15	32 c
Intron 6	13286	rs4646140	G/A	—	0	1
Intron 9	25082	NEW	G/A	—	0	1
Intron 10	25424	NEW	G/A	—	0	1
Intron 10	27418	rs4646165	G/A	—	0	1
Intron 12	28946	rs2301693	C/T	—	0	2
Intron 12	29018	rs2301692	A/G	—	0	2
Intron 14	30816	NEW	A/G	—	1	1
Intron 14	30867	rs4646174	C/G	—	0	2
Intron 14	33121	NEW	G/C	—	1	0
Exon 15	33205	NEW	A/G	N/S	0	1
Intron 16	36655	NEW	G/A	—	0	1
Intron17	38926	NEW	C/T	—	0	1
Exon 18 (3′‐UTR)	39663	NEW	C/G	—	0	1
Exon 18 (3′‐UTR)	39705	NEW	A/G	—	0	1
Exon 18 (3′‐UTR)	39844	NEW	G/A	—	3	4 c
					No. of samples tested = 20	No. of samples tested = 57

Position numbers indicate distance from 5′ end of the original exon 1.

Newly identified SNPs are shown as NEW.

Minor allele frequencies of the SNPs shown in bold and italic were higher than 0.05.

Genotype and Allele Frequency of Three SNPs

Two SNPs in intron 3 and exon 18 with minor allele frequencies higher than 0.05 and a newly identified non‐synonymous SNP, N638S in exon 15 were analyzed in all samples (Table IV). Relative positions of these SNPs are shown in Figure 1. Genotyping results by direct sequencing method were confirmed by RFLP or SSCP methods. Because ACE2 is located to the X chromosome in humans, samples from both males and females were analyzed, respectively. Two possible non‐synonymous SNPs that are shown in the dbSNP database (rs4646116 and rs11798104) were not found in our samples this time. When the antibody‐negative contacts group was compared with antibody‐positive group including SARS cases in either males or females, no difference was observed between the two groups both in regards to genotype and allele frequencies. Comparison between antibody‐positive contacts and SARS cases, and comparison between contacts and non‐contacts did not show any significant differences in genotype and allele frequencies of the tested polymorphisms.

Table IV

Genotype and Allele Distribution of Three Single Nucleotide Polymorphisms (SNPs)

				SARS cases	Contacts		Non‐contacts
					Antibody (+)	Antibody (−)
Intron 3 (rs2285666)
	Male	Genotype/allelea no. (frequency)	A	5 (0.38)	4 (0.57)	21 (0.54)	5 (0.31)
			G	8 (0.62)	3 (0.43)	18 (0.46)	11 (0.69)
			Total no.	13	7	39	16
	Female	Genotype no. (frequency)	A/A	12 (0.39)	4 (0.44)	15 (0.31)	11 (0.33)
			A/G	16 (0.51)	3 (0.33)	24 (0.50)	17 (0.52)
			G/G	3 (0.10)	2 (0.22)	9 (0.19)	5 (0.15)
			Total no.	31	9	48	33
		Allele no. (frequency)	A	40 (0.65)	11 (0.61)	54 (0.56)	39 (0.59)
			G	22 (0.35)	7 (0.39)	42 (0.44)	27 (0.41)
Exon 15 (N638S)
	Male	Genotype/allele no. (frequency)	A	13 (1.00)	7 (1.00)	39 (1.00)	17 (1.00)
			G	0 (0.00)	0 (0.00)	0 (0.00)	0 (0.00)
			Total no.	13	7	39	17
	Female	Genotype no. (frequency)	A/A	31 (1.00)	8 (0.89)	47 (0.98)	33 (1.00)
			A/G	0 (0.00)	1 (0.11)	1 (0.02)	0 (0.00)
			G/G	0 (0.00)	0 (0.00)	0 (0.00)	0 (0.00)
			Total no.	31	9	48	33
		Allele no. (frequency)	A	62 (1.00)	17 (0.94)	95 (0.99)	66 (1.00)
			G	0 (0.00)	1 (0.06)	1 (0.01)	0 (0.00)
Exon 18 (3′‐UTR)
	Male	Genotype/allele no. (frequency)	G	12 (0.92)	7 (1.00)	37 (0.95)	17 (1.00)
			A	1 (0.08)	0 (0.00)	2 (0.05)	0 (0.00)
			Total no.	13	7	39	17
	Female	Genotype no. (frequency)	G/G	27 (0.87)	8 (0.89)	46 (0.96)	29 (0.88)
			A/G	4 (0.13)	1 (0.11)	2 (0.04)	4 (0.12)
			A/A	0 (0.00)	0 (0.00)	0 (0.00)	0 (0.00)
			Total no.	31	9	48	33
		Allele no. (frequency)	G	58 (0.94)	17 (0.94)	94 (0.98)	62 (0.94)
			A	4 (0.06)	1 (0.06)	2 (0.02)	4 (0.06)

Genotype distribution is the same as allele distribution in male.

DISCUSSION

During the worldwide outbreak of SARS in 2003, a subset (about 20%–30%) of SARS patients required mechanical ventilation, having developed pneumonia. The fatality rate was 11%, although the majority of patients recovered without unfavorable outcome [Peiris et al., 2003b]. As a natural consequence, asymptomatic individuals produce antibodies against SARS‐CoV in their sera [Ip et al., 2004; Woo et al., 2004]. In one of the studies, it was shown that 2.3% of contacts who did not develop clinical SARS had serum antibody titer over the threshold [Ip et al., 2004], and this implies the presence of asymptomatic individuals.

We hypothesized that the functional polymorphism of ACE2, which is considered as being a virus receptor of SARS‐CoV, might influence the clinical history of SARS‐CoV infection at least in part. This is because, a variation of the co‐receptor to HIV, CCR5‐Δ32 where allele frequency is approximately 10% in the European population [Martinson et al., 1997], has been well known to resist HIV infection and alter its clinical course [Dean et al., 1996; Liu et al., 1996; Samson et al., 1996].

Using the PCR‐based cloning procedure, we identified for the first time an alternative exon upstream of the original exon 1 of ACE2 that is expressed in various organs, including the lung and trachea, primary‐cultured bronchial epithelial cells, and the small intestine. These are considered to be important replication sites of SARS‐CoV [Haagmans et al., 2004]. Both 5′‐and 3′‐ends of the intron between the new alternative exon and the original exon 1 followed the GT/AG rule of Breathnach and Chambon [1981]. Although the organ specificity of the transcripts was not confirmed in this study due to the limitation of non‐quantitative PCR amplification, implication of the new exon was definitely shown in the lung and small intestine. Also, we found the extended region of the original exon 1, 65 bp on the 5′ side. Neither the new alternative exon nor the new extended region of exon 1 gave rise to a new coding region and they were considered as 5′‐untranslated region.

It was recently reported that genetic variations of ACE2 did not affect SARS susceptibility or outcome in Hong Kong [Chiu et al., 2004]. In that study, five intronic SNPs (rs2106809, rs2285666, rs4646142, rs714205, and rs2074192) were chosen and analyzed in a case‐control manner, based on the previously known exon‐intron structure and SNPs already registered in the database. By contrast, we attempted to analyze not only previously known SNPs but also variations newly identified among actual SARS patients and contacts. Based on the information from the exon‐intron structure of ACE2 cloned by ourselves, we searched for nucleotide sequences in all the exons including the new alternative exon and the corresponding 5′‐flanking region, which are thought to contain promoters of the new exon and the original exon 1. We found one novel non‐synonymous substitution N638S and 18 non‐coding SNPs including two relatively common SNPs with minor allele frequency higher than 5%. We selected these SNPs and analyzed them furthermore in a case‐control manner, because, while they are rare occurrence, non‐synonymous substitution may directly modulate the function of the protein, and because relatively common SNPs can often be used as markers to ascertain a causative variation. Of 19 SNPs found in this study, 13 were new polymorphisms, 3 of which were located in 3′‐UTR. Two possible non‐synonymous SNPs in dbSNP database were not found in the population tested. Judging from the results so far obtained in this case‐control study, there was no statistical evidence that ACE2 polymorphisms affect SARS infection or alter its clinical course. However, type II error was not negligible because of a relatively small size of samples tested.

Taking also into consideration, the results from a previous study of ACE2 polymorphisms by others [Chiu et al., 2004], it is unlikely that the genetic defect of ACE2 is involved in the disease resistance that has been shown in CCR5‐Δ32 in HIV‐1 infection cases. Nevertheless, this newly identified alternative 5′‐untranslated exon expressed in the lung, and also newly recognized polymorphisms in this study might be of great help concerning investigations into the regulation of ACE2 gene expression and the possible significance of the variations in further more in‐depth studies.