ARTICLE IN PRESS

Microbiological Research 163 (2008) 414423

www.elsevier.de/micres

A type-1 metacaspase from Acanthamoeba

castellanii
Wendy C. Trzyna, Xavier D. Legras, John S. Cordingley
Department of Biological Sciences, Louisiana Tech University, PO Box 3179, Ruston, Louisiana, LA 71 272, USA
Received 28 April 2006; accepted 29 June 2006

KEYWORDS
Acanthamoeba
castellanii;
Type-1 metacaspase;
Encystation;
Signaling complex

Summary
The complete sequence of a type-1 metacaspase from Acanthamoeba castellanii is
reported comprising 478 amino acids. The metacaspase was recovered from an
expression library using sera specic for membrane components implicated in
stimulating encystation. A central domain of 155 amino acid residues contains the
Cys/His catalytic dyad and is the most conserved region containing at least 30 amino
acid identities in all metacaspases. The Acanthamoeba castellanii metacaspase has
the most proline-rich N-terminus so far reported in type-1 metacaspases with over 40
prolines in the rst 150 residues. AlaProPro is present 11 times. Phylogenies
constructed using only the conserved proteolytic domains or the complete sequences
show identical branching patterns, differing only in the rates of change.
& 2006 Elsevier GmbH. All rights reserved.

Introduction
Acanthamoeba castellanii is a small, free-living
amoeba although some isolates are opportunistic
human pathogens (Marciano-Cabral et al., 2000). It
is closely related to the cellular slime mold
Dictyostelium discoideum as shown by a number
of studies e.g. proteasome subunit sequence alignments (Bouzat et al., 2000). Acanthamoeba castellanii feeds phagocytically on bacteria and has
the ability to encyst, manufacturing a celluloseCorresponding author. Tel.: +318 257 4573;

fax: +318 257 4574.

E-mail address: wtrzyna@latech.edu (W.C. Trzyna).

containing cyst wall in order to survive periods of

adverse conditions. When conditions improve,
the amoeba emerges from the cyst and recommences growth. The environmental circumstances
that trigger encystation and the molecular details
of the pathways leading from the stimulus to the
mature cyst are incompletely understood. However, various specic conditions have been shown
to trigger encystation experimentally in rich media,
among which are increasing osmolarity (Cordingley
et al., 1996) and various monoclonal antibodies
(mAbs) that bind to the surface of the amoebae
(Villemez et al., 1985).
The mAbs that bound to the surface of the
amoeba and under specic circumstances triggered

0944-5013/$ - see front matter & 2006 Elsevier GmbH. All rights reserved.
doi:10.1016/j.micres.2006.06.017

ARTICLE IN PRESS
A type-1 metacaspase from Acanthamoeba castellanii
encystation potentially provided ways to identify
components of membrane complexes, possibly
proteins, involved in triggering encystation. However, upon further analysis many of the mAbs
binding to the surface recognized epitopes that
were either completely or partly composed of
carbohydrates. Since recombinant bacterial proteins are not glycosylated, the mAbs could not be
used to directly screen an expression library and
therefore an indirect approach was used to identify
protein components in these membrane complexes.
In this procedure we isolated membrane complexes, solubilized with non-ionic detergent in
order to preserve non-covalent associations, using
one of the mAbs that recognized a carbohydrate
determinant on the surface of the amoebae.
Proteins from these immune complexes were used
to raise polyclonal sera and using these sera we
isolated expression clones from a cDNA library. One
of the clones identied in this way encoded the
type-1 metacaspase reported here.
Metacaspases are a group of genes encoding
cysteine proteases found in higher plants, fungi and
protists. They have been classied into two distinct
subgroups, designated types 1 and 2 (Uren et al.,
2000). Type-1 metacaspases have diverse N-terminal extensions, often proline-rich, which are
absent in type-2 molecules and a central proteolytic domain. Type-2 metacaspases have an Nterminal proteolytic domain and C-terminal regions
varying in both length and sequence. There is a
single type-1 metacaspase gene in the Saccharomyces cerevisiae genome but, in contrast, there
are multiple metacaspase genes, both types 1 and
2, in multicellular plants such as Arabidopsis
thaliana in which the multiple gene copies are
probably paralogues arising by gene duplication and
subsequent divergence of the copies during the
evolution of the increasing cellular diversity in
higher plants (Vercammen et al., 2004).
D. discoideum has no metacaspase gene, however the D. discoideum genome has a paracaspase
gene that was used as the starting sequence for the
PSI-BLAST searches that led to the identication of
the metacaspase gene family (Uren et al., 2000).
The human paracaspase identied in the same
study by Uren et al. (2000) is related to the
caspases, the well-characterized group of cysteine
proteases that form an integral part of the
apoptotic machinery in metazoa (Cohen, 1997).
The distant but signicant homologies between
caspases, paracaspases and metacaspases have
prompted many efforts to establish functional
parallels between these groups of proteins. Metacaspases have since been shown to be involved in
apoptotic-like cell death in Saccharomyces

415
cerevisiae (Silva et al., 2005). However, they are
also implicated in other processes such as sporulation in Aspergillus nidulans (Thrane et al., 2004)
and plant embryogenesis (Suarez et al., 2004;
Bozhkov et al., 2005) in which specic cells in the
plant embryo undergo a form of programmed cell
death. These studies suggest that the diverse
intracellular roles of metacaspases are still incompletely understood.
We report here the identication of a type-1
metacaspase from Acanthamoeba castellanii and
report a comparative study with other type I
metacaspases and a discussion of possible functional roles.

Materials and methods

Acanthamoeba castellanii cultures
Acanthamoeba castellanii (Neff) cultures were
grown axenically as described previously except
glucose was included at a nal concentration of
1.5% or 3% w/v (Cordingley et al., 1996).

Immunoprecipitation with F1 Mab

Monoclonal antibody F1 (Villemez et al., 1985)
was linked to afgel and used for immunoprecipitations from an Acanthamoeba castellanii lysate.
1  108 Acanthamoeba castellanii cells were harvested and lysed in 50 mls 2% Np40/1  PBS/
protease inhibitor cocktail (4 mM pefabloc,
0.7 mg/ml pepstatin A, 5 mg/ml Leupeptin, 50 mg/
ml TLCK). F1-afgel or afgel alone was added to
lysate cleared by centrifugation at 10,000 rpm
for10 min and agitated gently at 4 1C for 2 h. The
F1-afgel was spun down, washed 3  with (0.2%
Np40/1  PBS/protease inhibitor cocktail) followed
by 3  with PBS. The F1-afgel with bound material
was boiled in SDSPAGE sample buffer, beads were
removed by centrifugation, and samples electrophoresed on 820% gradient gels as described
below. Comassiae blue-stained protein bands were
excised from gels and used to immunize mice as
described below.

SDSPAGE gels and western blots

The gels used in these experiments were 820%
gradient gels in the buffer conditions of Leammli
(Laemmli, 1970). Protein samples for gels were
boiled in sample buffer (10% SDS, 5% 2-mercaptoethanol, 125 mM TrisCl pH 6.8, 10% glycerol). For
western blots, samples were electrophoresed on

ARTICLE IN PRESS
416
SDSPAGE gels, and electro-blotted to PVDF (Immobilon-P, Millipore). Immunodetection was carried out using Western Breeze kits from Amersham,
according to the manufacturers instructions.

Immunizations
Bands were excised from coomassie blue stained
SDSPAGE gels as described previously (Trzyna and
Cordingley, 1993). Briey, gel slices were equilibrated in PBS and homogenized in RIBI adjuvant
(Corixa). Balb/c mice were immunized subcutaneously with 200 ml of homogenate (0.51.0 mg
protein/dose). Mice were boosted twice at 1014day intervals with an identical preparation and sera
collected 1014 days later.

Screening cDNA expression library

The Acanthamoeba castellanii cDNA library used
in these studies was kindly provided by Dr. Erik
Bateman (University of Vermont). Approximately
3  106 phage were plated on bacterial lawns (XL1Blue). Plates were overlaid with nitrocellulose
lters presoaked in IPTG (10 mM) and incubated
for 35 h. Filters were processed and screened with
the PicoBlue Immuno-screening kit (Stratagene).
Polyclonal sera were used at a 1:2000 dilution.
Positive plaques were picked and subjected to
secondary and tertiary screenings. Phagemids were
excised from positive clones according to Stratagene protocols and the inserts sequenced.

PCR and rapid amplication of cDNA ends

(50 RACE)
Total RNAs were extracted from Acanthamoeba
castellanii using Trizols according to manufacturers
instructions (Invitrogen) and further puried using
the RNeasys kit (Qiagen). Polyadenylated mRNAs
were isolated using Oligotexs Direct mRNA kit
(Qiagen). Puried poly A+ mRNAs or total RNA were
used in two rounds of 50 RACE (First Choice RLM-RACE
kit, Ambion) according to the manufacturers protocol with the following modications. The reverse
transcription step was carried out at 48 1C for 1 h
using either random hexamer primers or a metacaspase specic primer, GSP-3 (50 GGTCATCGTCAACAATCTGG30 ). Complementary DNA made from
random hexamer primed mRNA was used for nested
PCR using reagents supplied with the kit and Supertaq
Plus polymerase (Ambion). Reverse primers were
GSP-2 (50 CTTGCCGTCCTTCATCTTCT) as the outer
primer and GSP-3 as the inner primer and the forward
primers were the outer and inner adaptor primers

W.C. Trzyna et al.

supplied in the kit. PCR conditions were 4 min at
94 1C; then 30 cycles of: 30 s at 94 1C, 45 s at 62 1C,
2 min at 72 1C; then 7 min at 72 1C. RT primed with
GSP-3 was followed by nested PCR using HotstarTaqs
Master Mix kit (Qiagen) with Q solution (Qiagen).
Reverse primers were GSP-3 as the outer primer and
GSP-4 (50 CGGATCATGTTGGCCTTGG30 ) as the inner
primer and forward primers were the outer and inner
adaptor primers supplied with the kit. The PCR
conditions were as follows: 15 min at 94 1C; then 30
cycles of: 30 s at 94 1C, 45 s at 58 1C, 1 min at 72 1C;
then 7 min at 72 1C. PCR products were gel puried
(QIAEXs II Gel Extraction Kit, Qiagen) and cloned into
TA pCRII (Invitrogen) according to the manufacturers
protocol. Full-length metacaspase was amplied from
rst strand cDNA using the primer set (forward:
50 AGAAGAAGAGAGCAGCAA30 , reverse: 50 TATGATCCAACCAACCCC30 ) and the HotStarTaq Master kit
(Qiagen) with the following conditions 15 min at
94 1C; then 35 cycles: 45 s at 94 1C, 45 s at 51 1C, 2 min
at 72 1C; then 10 min at 72 1C. PCR products were
cloned into TOPO pCR 2.1 (Invitrogen) and sequenced.

DNA sequencing and sequence analysis

DNA sequencing was carried out by the Davis
Sequencing facility, (Davis Sequencing, LLC, 1490
Drew Avenue, Suite 170, Davis, CA 95 616, USA).
Automated DNA sequencing results were analyzed
using the Omiga 2.0 sequence analysis package
from GCG and the Staden package of sequence
analysis programs available from http://www.mrclmb.ac.uk/pubseq/staden_home.html. The codonpreference analyses were performed using the
program Spin from the Staden package. Protein
alignments were constructed using Clustal-W
ver.1.6 and dendrograms from the alignments were
displayed as phylograms using the Treeview software package (Page, 1996).

Southern and Northern blots

Southern blots were carried out using standard
protocols and as described previously (Nene et al.,
1986). Northern blots were carried out according to
standard protocols. All lanes were loaded with
equal quantities of total RNA (20 mg/lane).

Results
One of the mAbs that bound to the surface of
Acanthamoeba castellanii and triggered encystation, referred to as F1 (Villemez et al., 1985),

ARTICLE IN PRESS
A type-1 metacaspase from Acanthamoeba castellanii
did not recognize mRNA in vitro translation
products but it did recognize native molecules
from logarithmically growing trophozoites on western blots. When these western blots were treated
using enzymes that specically degrade carbohydrates found on glycoproteins, (Glycopro Deglycosylation kit (Prozyme) according to protocols
provided with the kit) the binding of the F1 mAb
was signicantly reduced (data not shown). These
two observations taken together suggested that the
F1 mAb probably recognized a carbohydrate-containing epitope on the surface of the trophozoite,
possibly a glycoprotein.
Complexes were immunoprecipitated by the
F1 mAb from membranes solubilized using non-ionic
detergents. These immunoprecipitates were fractionated by SDSPAGE and a number of protein
bands were reproducibly observed by coomassie
blue staining in multiple independent preparations.
Coomassie blue staining bands from these polyacrylamide gels were excised and used to raise
polyclonal hyperimmune sera. Using these polyclonal antisera we screened an Acanthamoeba
castellanii cDNA expression library and identied
specic clones. One of the cDNA clones (F102b)
identied in this way encoded the 30 -end of a
polyadenylated metacaspase mRNA.
The sequence of the clone (Genbank Acc. No.
AF480890) was used to search the available
databases. When the initial cDNA clone was
isolated the most signicant match in the databases
at that time to an annotated sequence was to a
gene identied as a metacaspase from the Schizosaccharomyces pombe genome (Acc. No. 7491859).
The S. pombe metacaspase was a type 1 but since
F102b was incomplete at the 50 end we initially
could not determine whether F102b was a type 1 or
2 molecule. Consequently primers based on the
initial sequence were used to extend the sequence
by 50 RACE. By combining the results of two rounds
of RACE cloning and sequencing, the full-length
sequence reported here was compiled. The increasing length of the cDNA sequence is reected in
successive updates of the Genbank submission
(AF480890) leading to the present full-length
sequence. To conrm the complete sequence,
primers in the 30 and 50 untranslated regions were
used to amplify a complete copy of the openreading frame by RT-PCR using mRNA from Acanthamoeba castellanii. The resulting PCR product was
cloned and sequenced in both directions. The
sequence conrmed the Genbank submission exactly, encoding a type-1 metacaspase comprising
478 amino acids (Fig. 1).
After isolating the initial cDNA clone genomic
clones were isolated from a library (kindly provided

417
by Dr. E. Bateman) and sequenced with the
intention of completing the metacaspase sequence
from genomic clones. However the identied
genomic clones, while they did agree with and
extend the cDNA clones, were interrupted by
multiple short introns and led us to the conclusion
that in order to be sure the deduced sequence was
expressed it would be preferable to isolate cDNA
clones covering the complete sequence. The fulllength sequence was entirely derived from cDNA
prepared from polyadenylated message and consequently represents expressed sequences. This is
particularly relevant for the type-1 metacaspase
since the translated sequence contains a section of
non-conserved sequence that is probably processed
out of the active protease (region 2, Fig. 2). To be
sure that this region was present in the mature
mRNA and not, for example, removed by splicing of
the precursor RNA we prepared and sequenced the
clones as described.
At the time of writing, there were nearly 100
annotated metacaspase protein sequences in the
NCBI sequence databases. However, for the sake of
clarity, a selected subset of these sequences is
presented in Fig. 1 aligned with the Acanthamoeba
castellanii metacaspase. The sequence of the
mRNA from which the protein sequence was
deduced is not shown in Fig. 1 but is present in
full in the Genbank submission (AF480890). All of
the computer analyses presented here used amino
acid sequences. Figure 2 shows a Dot-Matrix
analysis of 5 metacaspases, one of which is the
Acanthamoeba castellanii metacaspase, compared
to the Saccharomyces cerevisiae metacaspase
sequence. The Saccharomyces cerevisiae sequence
was used as the comparator sequence in this dotplot since it is the single metacaspase gene from a
fully sequenced genome in contrast to the other
species whose genomes are less thoroughly nished. Metacaspases were selected to cover the
phylogenetic groups possessing metacaspase genes.
Kluyveromyces lactis was chosen as a yeast species
closely related to Saccharomyces cerevisiae, Aspergillus nidulans is an ascomycete fungus,
Acanthamoeba castellanii is the subject of this
paper and a representative of the crown group
(Philippe et al., 2000), Arabidopsis thaliana is an
angiosperm and Trypanosoma brucei is a kinetoplastid protist that branches very early in the
eukaryotic line of descent. Due to the relatively
low level of match outside the proteolytic domain
the T. brucei sequence is omitted from the
alignment in Fig. 1.
The type-1 metacaspase genes are divisible into
four conceptually distinct regions indicated in
Fig. 2. Region 1 is the region of 153155 residues

ARTICLE IN PRESS
418

W.C. Trzyna et al.

At

1 . . . . . . . . 10
- - - - - - - - - -

. . . . . . . . . 20
- - - M L L L V D C

. . . . . . . . . 30
S S C R T P - - L H

. . . . . . . . . 40
L P P G - - - - A T

. . . . . . . . . 50
R I R C A I C H A F

. . . . . . . . . 60
T L I A P - - - - -

Ac

M A Y P Y G A P G G

G Y G V P P P D P Y

Q Q G G Q P G Y G A

P P P G - - - - - Q

G Y G M G P P Q G M

G Y G A P P - - Q G

An

- - - - - - - - - -

- - - - - - - M H H

H H Q Q P S - Y G S

G Y P G - - - - - Q

A Y R - - - - - - -

Q Q Q N P - - Y P Q

Sc

- - - - - - - - - -

- M S L E V Y L N Y

H Q R R P T R F T I

M Y P G S G R Y T Y

N N A G G N N G Y Q

R P M A P P P N Q Q

At

. . . . . . . . . 70
- E P R - L Q S H A

. . . . . . . . . 80
S A S P F P F P N S

. . . . . . . . . 90
S - - - - - - P A P

. . . . . . . . . 100 . . . . . . . . . 110 . . . . . . . . . 120

S T F I Y P P P T P
S P Y T H A P - - - - - H A P S P F N

Ac

Y G A P P P Q Q G Y

D E R W M Q A P S M

G A P P Q G M G M G

Y S A P P M G A P P

A G Y G A P P Q G M

G Y G A P P A G Y G

An

Y G H P S P Q P Y P

P Q N G Y S H P S S

G - - - - - - Y P P

S P A P P N G G Q M

Y H G R Q P S - - -

- - - Y P P N Q Y P

Sc

Y G Q Q Y G Q Q Y E

Q Q Y G Q Q Y G Q Q

N D Q Q - - - F S Q

Q Y A P P P G P P P

M A Y N R P V - - -

- - - Y P P P Q F Q

At

. . . . . . . . . 130 . . . . . . . . . 140 . . . . . . . . . 150 . . . . . . . . . 160 . . . . . . . . . 170 . . . . . . . . . 180

H A P - - - - - - - - - - - - - - - - P D S Y P F T H A
P P A S S P F N H A
P P G - - - - P P P
P - - V H G Q K R A

Ac

A P P Q G M G P P Q

G Y G A P P Q G Y G

Q M G Y P P Q G G A

P P P G Q Q M M G Y

G G G G G F T Q F Q

P S A C T G R K K A

An

P - - - - - - - - -

- - - - - - - - - -

- A H G G P T A P P

T N P Q A F G H G A

P Q G - - - Y N F Q

Y S R C T G K R K A

Sc

Q E Q A K A Q L S N

G Y N N P N V N A S

N M Y G P P Q N M S

L P P P Q T Q T I Q

G T D - - - Q P Y Q

Y S Q C T G R R K A

At

. . . . . . . . . 190 . . . . . . . . . 200 . . . . . . . . . 210 . . . . . . . . . 220 . . . . . . . . . 230 . . . . . . . . . 240

V I V G V S Y K N T
K D E L K G C I N D
A N C M K F M L M K
R F Q F P - - E S C
I L M L T E E E A D
P M R W P T K N N I

Ac

L L I G I N Y V N S

Q R P L K G C V N D

V Q N I R R F I T Q

R F G F R D A P D S

M I V L T D D Q N D

P S R R P T K A N M

An

L L I G I N Y F G Q

K G Q L R G C I N D

V K N M S T Y L N Q

N F G Y A - - R E D

M V I L T D D Q Q N

P M S Q P T K A N I

Sc

L I I G I N Y I G S

K N Q L R G C I N D

A H N I F N F L T N

G Y G Y S - - S D D

I V I L T D D Q N D

L V R V P T R A N M

At

. . . . . . . . . 250 . . . . . . . . . 260 . . . . . . . . . 270 . . . . . . . . . 280 . . . . . . . . . 290 . . . . . . . . . 300

T M A M H W L V L S
C K P G D S L V F H
F S G H G N N Q M D
D N G D E V D G F D
E T L L P V D H R T
S G V I V D D E I N

Ac

I R A M Q W L I Q G

A Q P G D S L F L H

F S G H G G Q V R D

T D G D E D D G F D

E T I L P E D Y A S

A G Q I V D D D L H

An

L R A M H W L V K D

A Q P N D S L F F H

Y S G H G G Q T P D

L D G D E D D G Y D

E V I Y P V D F R V

A G H I V D D E M H

Sc

I R A M Q W L V K D

A Q P N D S L F L H

Y S G H G G Q T E D

L D G D E E D G M D

D V I Y P V D F E T

Q G P I I D D E M H

At

. . . . . . . . . 310 . . . . . . . . . 320 . . . . . . . . . 330 . . . . . . . . . 340 . . . . . . . . . 350 . . . . . . . . . 360

A T I V R P L P Y G
V K L H A I V D A C
H S G T V M D L P Y
L C R M D R L G N Y
E W E D H R P K T G
M W K G T S G G E V

Ac

K I L V K H L P P G

V R L T V V F D S C

H S G T A M D L P Y

V Y N E N G C I D S

P S M A S G K K A K

K L Q K K M M K K Q

An

R I M V K P L Q P G

V R L T A I F D S C

H S G S A L D L P Y

I Y S T Q G I L K E

P - - - - - N L A K

E A G Q G L L G V I

Sc

D I M V K P L Q Q G

V R L T A L F D S C

H S G T V L D L P Y

T Y S T K G I I K E

P - - - - - N I W K

D V G Q D G L Q A A

At

. . . . . . . . . 370 . . . . . . . . . 380 . . . . . . . . . 390 . . . . . . . . . 400 . . . . . . . . . 410 . . . . . . . . . 420

F S F T G C D D D Q
T S A D T P Q L S G
S A W T G A M T - Y
A F I Q A I E R G H
G M T Y G S L L N A
M R S T V H E I F D

Ac

K K M K D G K G K G

K G K G K G K H Q Y

A A T G G A G - - G

P T L S G Q A A K Q

T M A D V I M F S G

C R D S Q T A A D T

An

S S Y A R G D M G G

M M S T A V G F L K

K A A K G D - - - E

A Y Q R T K Q T K T

S P A D V I M W S G

S K D D Q T S Q D A

Sc

I S Y A T G N R A A

L I G S L G S I F K

T V K G G M G N N V

D R E R V R Q I K F

S A A D V V M L S G

S K D N Q T S A D A

At

. . . . . . . . . 430 . . . . . . . . . 440 . . . . . . . . . 450 . . . . . . . . . 460 . . . . . . . . . 470 . . . . . . . . . 480

K N K G R E L V E V
G G A D F L S T L L
G L L I L G A S P P
D E E E E V N Q A P
- - - - - Q K T Q E
P Q L S A N E A F A

Ac

S F A G M G Q T G A

M S Y A L L T V L G

K T P K L S Y H D L

L Q Q M R Y L L N S

G Q G G R T F T Q V

P Q L S T G R P M D

An

Q I A G Q - A T G A

M S W A F I T A M R

K N P Q Q S Y V Q L

L N S I R D E L S T

- - - - - R Y T Q K

P Q L S S S H P L D

Sc

V E D G Q - N T G A

M S H A F I K V M T

L Q P Q Q S Y L S L

L Q N M R K E L A G

- - - - - K Y S Q K

P Q L S S S H P I D

At

. . . . . . . . . 490 . . . . . . . . . 500 . . . . . . . . . 510 . . . . . . . . . 520 . . . . . . . . . 530 . . . . . . . . . 540

V Y E K P F S L

Ac

M N A Q F I M

An

V N L L Y V M

Sc

V N L Q F I M

Figure 1. The Acanthamoeba castellanii type-1 metacaspase protein sequence aligned with three other type-1
metacaspase proteins. The metacaspase from Arabidopsis thaliana, sequence Atmc2 (AAP84707) [21] is designated At;
the metacaspase from Acanthamoeba castellanii (AF480890), Ac; from Aspergillus nidulans (genbank:AF528964), An,
and from Saccharomyces cerevisiae (SCYOR197W); Sc. The numbering reects the overall length of the alignment not
the length of any one individual metacaspase. The alignment is broken into blocks of ten residues. Amino acid identities
are emphasized in dark gray, conserved residues in light gray.

ARTICLE IN PRESS
A type-1 metacaspase from Acanthamoeba castellanii

not conserved. Region 3 is the N-terminal domain

that also diverges rapidly in sequence when
compared to region 1 but also shows the codon
preferences of the host. Region 4 is the C-terminal
domain, which aligns to the C-terminal residue in
all type-1 metacaspases with the exception of the
T. brucei sequences, which have between 11 and 27
additional amino acids at the C-termini. The block
of six residues, QXPQLS, beginning with the
glutamine at position 469 in Fig. 1, is identical in
all the type-1 metacaspases with the exception of
T. brucei which has conservative substitutions. The
C-terminal extensions in the T. brucei sequences
are clear when these conserved hexamers are
aligned (not shown).
The N-terminal domains of type-1 metacaspases
(region 3, Fig. 2) share some sequence features but
as Figs. 1 and 2 illustrate they are signicantly
more divergent than the conserved proteolytic
domains (region 1). The N-terminal domains (region
3) are proline rich and internally repetitive but
show little long-range sequence conservation and
their degree of proline enrichment varies markedly
between species. The proline-containing motifs are
tabulated in Table 1 to illustrate the diversity of the
proline-containing motifs in this region. Regions
1+2+4 comprise an almost constant number of
proline residues (1373), whereas region 3 contains
from 4 prolines in T. brucei to 40 proline residues in
the Acanthamoeba castellanii metacaspase. In
each case the Acanthamoeba castellanii sequence
show the largest number of each proline-containing
motif. Region 3 in T. brucei is not really a prolinerich region at all having a lower proportion of
prolines than the remainder of the molecule and all
other type-1 metacaspase N-termini. The motif
AlaProPro is the most numerous proline-rich
motif in Acanthamoeba castellanii. (Since the
acronym APP has been widely used to denote
Amyloid precursor protein in many studies targeting
Alzheimers disease, the three-letter abbreviations
(AlaProPro) have been used throughout to avoid
any confusion).

encoding the proteolytic domain with the CysHis

catalytic dyad and comprises the most conserved
residues in all type-1 metacaspases. Region 2 is a
segment that is probably processed out of the
mature protein by analogy with caspase activation
and similar examples. Region 2 is probably translated since the codons used in region 2 reect the
codon preferences of the host (Staden and McLachlan, 1982) but the amino acid sequence itself is

Figure 2. A dot plot of ve full-length type-1 metacaspase proteins compared to the single type-1 metacaspase
from Saccharomyces cerevisiae (WSCYOR197W). Full
length amino acid sequences of type-1 metacaspase
proteins from K. lactis (Kl)(Aspergillus nidulans
(An)(AF528964) Acanthamoeba castellanii (Ac)(AF480890),
A. thaliana (At)(AAP84707) and Trypanosoma brucei
(Tb)(AJ437301) were joined into one long composite
sequence of 2109 amino acids and compared in a
Dotplot analysis with the single type-1 metacaspase
from Saccharomyces cerevisiae. The dot plot used a
window of 12 amino acids, a stringency of 30%, hash size
of 2 and the BLOSUM 45 scoring matrix using OMIGA
software from GCG. Region 1, conserved proteolytic
domain; Region 2, processed domain; Region 3, Nterminal domain; Region 4 conserved C-terminal domain.

Table 1.

419

Proline containing motifs in ve metacaspases

Motif regions

T. brucei4
A. thaliana2
A. castellanii
A. nidulans
K. lactis
S. cereviseae

Pro

ProPro

ProProPro

AlaProPro

1+2+4

1+2+4

1+2+4

1+2+4

4
30
40
25
19
23

12
15
14
13
11
10

0
9
19
6
6
11

0
1
1
0
0
0

0
3
4
0
2
5

0
0
0
0
0
0

0
3
11
2
1
2

0
0
1
0
0
0

ARTICLE IN PRESS
420
Two phylograms are presented in Fig. 3. Phylogram 1 was constructed using Clustal alignments of
the full-length metacaspase genes with the trees
displayed using Treeview (Page, 1996) with the
early-branching T. brucei sequence used as the outgroup. Phylogram 2 in Fig. 3 was constructed from
an alignment of only region 1. The scales are
identical in both phylograms and the longer
branches in Phylogram 1 reect the greater
diversity when the N-termini are included in the
alignment; however, the pattern and branching
order observable in the two phylograms are
identical. When the N-terminal regions are aligned
independently the resulting tree has branches so
deep that the branching order has no real statistical signicance. The tree structure seen for the
complete metacaspase sequences is clearly the
result of the alignment of the conserved regions,
whereas the N-termini show no strong sequence
relationships.
Figure 4 shows Northern and Southern blots
probed with the metacaspase cDNA clone insert
(F102b). Message is present during growth and
appears to increase at the onset of stationary phase
when the trophozoites are initiating encystation.
Preliminary data acquired from real-time RT-PCR
also support this trend (not shown). The Southern
blot shows that the gene is present in the genome
either once or at the most a few times. The pending
complete genome sequence of Acanthamoeba
castellanii
[http://www.genome.gov/13014443]
will resolve the question of the exact gene number.

W.C. Trzyna et al.

Figure 4. Southern and Northern blots of Acanthamoeba

castellanii material probed with the F102b metacaspase
cDNA clone insert. The left panel is a genomic Southern
blot, the arrows indicate Molecular Weight markers, 10,
6, 4, 2, and 1 kb from the top down. The right hand panel
is a Northern blot using total RNA from; lane 1,log phase
(2  106 cells/ml); lane 2, stationary phase (8  106 cells/
ml); lane 3, 5 h stationary). Equal amounts of RNA were
loaded from each time point.

Discussion

Figure 3. Phylogram based on six type-1 metacaspase

amino acid sequences. The six metacaspase sequences
are the same ones used to construct the dot plot in Fig. 2,
Saccharomyces cerevisiae, K. lactis, Aspergillus nidulans,
Acanthamoeba castellanii, Arabidopsis thaliana and T.
brucei. The phylograms were constructed using Clustal-W
ver 1.6 and the alignments displayed and scaled using
Treeview software [9]. The scale bar indicates 0.1
substitutions per residue. The outgroup was set as T.
brucei.

The Acanthamoeba castellanii metacaspase gene

described here was isolated from an expression
library using antisera raised against a proteincontaining band from SDSPAGE of immune complexes precipitated by a monoclonal antibody (F1)
that stimulates encystation. Isolating a metacaspase gene by an indirect method such as this is
clearly not conclusive proof that the metacaspase
is in the direct line of events leading to encystation. However, the mAbs did provide a very
effective tool to focus our attention on a subset
of the cells proteins associated with the stimulation of encystation, one of which turned out to be
the metacaspase described here. Notwithstanding
the indirect nature of the isolation procedure, the
metacaspase is an inherently interesting sequence.
Alignment of the full-length protein sequences
with the Acanthamoeba castellanii type-1 metacaspase in Fig. 1 shows that region 1 (dened in
Fig. 2) is readily aligned with very few insertions
being necessary. Of the residues, 47% are identical

ARTICLE IN PRESS
A type-1 metacaspase from Acanthamoeba castellanii
in region 1 of the metacaspases from the four
species, Acanthamoeba castellanii, Aspergillus
nidulans, Saccharomyces cerevisiae and K. lactis.
Inclusion of the two metacaspase sequences from T.
brucei and Arabidopsis thaliana reduces the number of identities in region 1 to approximately 26%
reecting their more distant evolutionary relationships. The conserved block QXPQLS which is found
in all the metacaspases presented in Fig. 1,
including the Arabidopsis thaliana type-1 metacaspases but conservatively substituted in T. brucei
which is consistent with the early divergence of this
species. This is reected in Fig. 2 and in the
phylograms in Fig. 3 where the early branching T.
brucei has been used as the out-group. There are
multiple type-1 metacaspases in both Arabidopsis
thaliana and T. brucei and only one specic
member was chosen for the gures. Inclusion of
all of the metacaspases from each of these species
results in an alignment and tree which root all of
the metacaspases from the same species together
and consequently one specic gene was used in the
gures and the calculated numbers apply to this
one example although the pattern is the same
for all the type-1 metacaspases from each of these
species.
The N-termini of all the metacaspases are the
most divergent regions. When the Acanthamoeba
castellanii N-terminal region (Fig. 2, region 3) is
used as the query sequence in a Blast search
(omitting the low complexity lter and using
composition based statistics) it does not nd
primarily other type-1 metacaspases. These
searches return a variety of matches from diverse
sources. For example, the most signicant match is
with a hypothetical protein-conserved from T.
brucei (gbAAZ13155.1), followed by a hypothetical protein from Cryptococcus neoformans
(gbAAW43431.1). Further down the signicant hits
list is a match from the echinoderm Strongylocentrotus purpuratus. However, when any of these
sequences are compared with the Acanthamoeba
castellanii full-length metacaspase there is no
similarity of any kind in regions 1, 2, and 4 (Fig.
2). The regions of identity include the sequence
GYGAPP, a motif that occurs in many proteins from
diverse species with no recent common ancestry.
For example, in the mouse protein (unnamed
protein product, gi12854364), the amino acid
sequence GYGAPPAGYGAPP is found exactly. These
similarities probably reect convergent evolution
in which the same sequence may have arisen
independently. The sequence in Acanthamoeba
castellanii is clearly an Acanthamoeba castellanii
translated sequence because of its codon preferences but sequence comparisons such as these

421
provide little information illuminating the function
or functions of these regions.
The type-1 metacaspase in Saccharomyces cerevisiae has been implicated in apoptosis-like cell
death induced by several environmental insults
including hyperosmolarity (Silva et al., 2005).
However increasing osmolarity triggers encystation
in Acanthamoeba castellanii rather than apoptosis
as in Saccharomyces cerevisiae (Cordingley et al.,
1996). Also of note, the related paracaspase gene
of D. discoideum is not involved in developmental
cell death in this organism (Roisin-Bouffay et al.,
2004) emphasizing that all members of the caspaselike families of genes are not necessarily involved in
cell death.
The intact F1 mAb (Yang and Villemez, 1994;
Villemez et al., 1985) (an IgG1) triggered encystation whereas monovalent Fab fragments did not
suggesting that cross-linking of the epitopes was
necessary to trigger encystation. In a similar way,
dimerization or trimerization of initiator caspases
has been linked to their activation (Muzio et al.,
1998). Long N-terminal extensions are a common
feature of these initiator or upstream caspases
as homotypic interaction domains involved in the
early stages of the signaling cascades leading to
apoptosis (Kischkel et al., 1995, Muzio et al., 1998,
Srinivasula et al., 1998). We suggest by direct
analogy that the N-terminal extension of type-1
metacaspases may be responsible for proteinprotein interactions initiating a signaling cascade
leading to encystation in Acanthamoeba castellanii
or apoptosis-like cell death in Saccharomyces
cerevisiae (Silva et al., 2005). The proline-rich
N-terminal domain may participate in proteinprotein interactions within the cell since polyproline motifs are established ligands for WW
domain proteins involved in intracellular signaling
networks (Kato et al., 2004; Ingham et al., 2005).
Further experiments are required to explore these
suggestions.
The increase in signal on the Northern blots
during the onset of stationary phase is striking and
has been supported by preliminary real-time RTPCR analysis of the mRNA levels during the onset of
encystation (data not shown). A similar increased
expression of metacaspase is seen during the
hyperosmolarity induced apoptosis-like cell death
in Saccharomyces cerevisiae (Silva et al., 2005) but
in yeast it accompanies cell death whereas in
Acanthamoeba castellanii it occurs in cells that are
beginning to encyst. Further experiments are
needed to investigate the involvement of the
metacaspase in this process.
Metacaspases, paracaspases and caspases are all
groups of cysteine proteases sharing the same

ARTICLE IN PRESS
422
evolutionary roots (Boyce et al., 2004). However,
caspases are proteases targeting aspartate residues
at the P1 position of the cleaved bond whereas the
type-2 metacaspases characterized to date target a
basic residue at this position (Watanabe and Lam,
2005, Bozhkov et al., 2005). In the gymnosperm
Picea abies, a metacaspase is involved with
programmed cell death during embryogenesis
(Bozhkov et al., 2005) showing that the metacaspase in P. abies is specic for substrates with
arginine or another basic amino acid at the P1
position. The P. abies metacaspase is a type-2
metacaspase as are the lysine or arginine preferring
type-2 metacaspases in Arabidopsis thaliana (Vercammen et al., 2004). The Acanthamoeba castellanii metacaspase described here is a type-1
enzyme and its specicity is uncharacterized.
However, whereas aspartic acids are found at the
processing sites in caspases we might anticipate
basic residues at processing sites in metacaspases.
The multiple basic residues between region 3 and
region 1 (RKK, residues 178180 in Fig. 1) are one
potential processing site as are the multiple basic
residues in region 2 of the Acanthamoeba castellanii sequence (poly-Ks and GK repeats, residues
320378 in Fig. 1). Further studies will resolve
these questions.
Considering these data implicating metacaspases
in either apoptosis and/or encystation in different
systems we suggest that these two processes may
share evolutionary roots and that apoptosis and
encystation may therefore be genuine homologs.
For a unicellular organism capable of encystation
such as Acanthamoeba castellanii the formation of
a cyst would seem to be a perfectly satisfactory
alternative to apoptosis, i.e. self-induced cell
death.

Acknowledgment
We thank Dr. Eric Bateman for DNA libraries.

References
Bouzat, J.L., McNeil, L.K., Robertson, H.M., Solter, L.F.,
Nixon, J.E., Beever, J.E., Gaskins, H.R., Olsen, G.,
Subramaniam, S., Sogin, M.L., Lewin, H.A., 2000.
Phylogenomic analysis of the alpha proteasome gene
family from early-diverging eukaryotes. J. Mol. Evol.
51 (6), 532543.
Boyce, M., Degterev, A., Yuan, J., 2004. Caspases: an
ancient cellular sword of Damocles. Cell Death Differ.
11 (1), 2937.

W.C. Trzyna et al.

Bozhkov, P.V., Suarez, M.F., Filonova, L.H., Daniel, G.,
Zamyatnin Jr., A.A., Rodriguez-Nieto, S., Zhivotovsky,
B., Smertenko, A., 2005. Cysteine protease mcII-Pa
executes programmed cell death during plant embryogenesis. Proc. Natl. Acad. Sci. USA 102, 1446314468.
Cohen, G.M., 1997. Caspases: the executioners of
apoptosis. Biochem. J. 326, 116.
Cordingley, J.S., Wills, R.A., Villemez, C.L., 1996.
Osmolarity is an independent trigger of Acanthamoeba
castellanii differentiation. J. Cell. Biochem. 61,
167171.
Ingham, R.J., Colwill, K., Howard, C., Dettwiler, S., Lim,
C.S., Yu, J., Hersi, K., Raaijmakers, J., Gish, G.,
Mbamalu, G., Taylor, L., Yeung, B., Vassilovski, G.,
Amin, M., Chen, F., Matskova, L., Winberg, G.,
Ernberg, I., Linding, R., Odonnell, P., Starostine, A.,
Keller, W., Metalnikov, P., Stark, C., Pawson, T., 2005.
WW domains provide a platform for the assembly of
multiprotein networks. Mol. Cell. Biol. 16, 70927106.
Kato, Y., Nagata, K., Takahashi, M., Lian, L., Herrero,
J.J., Sudol, M., Tanokura, M., 2004. Common mechanism of ligand recognition by group II/III WW domains:
redening their functional classication. J. Biol.
Chem. 279, 3183331841.
Kischkel, F.C., Hellbardt, S., Behrmann, I., Germer, M.,
Pawlita, M., Krammer, P.H., Peter, ME., 1995. Cytotoxicity-dependent APO-1 (Fas/CD95)-associated proteins form a death-inducing signaling complex (DISC)
with the receptor. EMBO J. 14, 55795588.
Laemmli, U.K., 1970. Cleavage of structural proteins
during the assembly of the head of bacteriophage T4.
Nature 227, 680685.
Marciano-Cabral, F., Puffenbarger, R., Cabral, G.A.,
2000. The increasing importance of Acanthamoeba
infections. J. Eukaryot. Microbiol. 47, 2936.
Muzio, M., Stockwell, B.R., Stennicke, H.R., Salvesen,
G.S., Dixit, V.M., 1998. An induced proximity model
for caspase-8 activation. J. Biol. Chem. 273,
29262930.
Nene, V., Dunne, D.W., Johnson, K.S., Taylor, D.W.,
Cordingley, J.S., 1986. Sequence and expression of a
major egg antigen from Schistosoma mansoni. Homologies to heat shock proteins and alpha-crystallins.
Mol. Biochem. Parasitol. 21, 179188.
Page, R.D.M., 1996. Treeview: an application to display
phylogenetic trees on personal computers. Comput.
Appl. Biosci. 12, 357358.
Philippe, H., Germot, A., Moreira, D., 2000. The new
phylogeny of eukaryotes. Curr. Opin. Genet. Dev 10,
596601.
Roisin-Bouffay, C., Luciani, M.F., Klein, G., Levraud, J.P.,
Adam, M., Golstein, P., 2004. Developmental cell
death in Dictyostelium does not require paracaspase.
J. Biol. Chem. 279 (12), 1148911494 Epub 2003
December 17.
Silva, R.D., Sotoca, R., Johansson, B., Ludovico, P.,
Sansonetty, F., Silva, M.T., Peinado, J.M., Corte-Real,
M., 2005. Hyperosmotic stress induces metacaspaseand mitochondria-dependent apoptosis in Saccharomyces cerevisiae. Mol. Microbiol. 58 (3), 824834.

ARTICLE IN PRESS
A type-1 metacaspase from Acanthamoeba castellanii
Srinivasula, S.M., Ahmad, M., MacFarlane, M., Luo, Z.,
Huang, Z., Fernandes-Alnemri, T., Alnemri, E.S.,
1998. Generation of constitutively active recombinant
caspases-3 and -6 by rearrangement of their subunits.
J. Biol. Chem. 273, 1010710111.
Staden, R., McLachlan, A.D., 1982. Codon preference and
its use in identifying protein coding regions in long
DNA sequences. Nucl. Acids Res. 10, 141156.
Suarez, M.F., Filonova, L.H., Smertenko, A., Savenkov,
E.I., Clapham, D.H., von Arnold, S., Zhivotovsky, B.,
Bozhkov, P.V., 2004. Metacaspase-dependent programmed cell death is essential for plant embryogenesis. Curr. Biol. 14, 339340.
Thrane, C., Kaufmann, U., Stummann, B.M., Olsson, S.,
2004. Activation of caspase-like activity and poly (ADPribose) polymerase degradation during sporulation in
Aspergillus nidulans. Fungal Genet. Biol. 41, 361368.
Trzyna, W.C., Cordingley, J.S., 1993. Schistosoma mansoni: genetic non-response to p40, the major antigen
of the egg, reveals a novel mechanism enhancing IgM
production during infection. Parasite Immunol. 15,
601611.

423
Uren, A.G., ORourke, K., Aravind, L.A., Pisabarro, M.T.,
Seshagiri, S., Koonin, E.V., Dixit, V.M., 2000. Identication of paracaspases and metacaspases: two ancient
families of caspase-like proteins, one of which plays a
key role in MALT lymphoma. Mol. Cell. 6, 961967.
Vercammen, D., van de Cotte, B., De Jaeger, G.,
Eeckhout, D., Casteels, P., Vandepoele, K., Vandenberghe, I., Van Beeumen, J., Inze, D., Van Breusegem,
F., 2004. Type II metacaspases Atmc4 and Atmc9 of
Arabidopsis thaliana cleave substrates after arginine
and lysine. J. Biol. Chem. 279, 4532945336.
Villemez, C.L., Carlo, P.L., Russell, M.A., 1985. Differentiation in Acanthamoeba castellanii is induced by
specic monoclonal antibodies. J. Cell. Biochem. 29,
373379.
Watanabe, N., Lam, E., 2005. Two Arabidopsis metacaspases AtMCP1b and AtMCP2b are arginine/lysinespecic cysteine proteases and activate apoptosis-like
cell death in yeast. J. Biol. Chem. 280, 1469114699.
Yang, S., Villemez, C., 1994. Cell surface control of
differentiation in Acanthamoeba. J. Cell. Biochem.
56, 592596.