Stage specific expression of poly(malic acid)-affiliated
genes in the life cycle of Physarum polycephalum
Spherulin 3b and polymalatase
Nadthanan Pinchai, Bong-Seop Lee and Eggehard Holler
Institut fu
¨
r Biophysik und Physikalische Biochemie der Universita
¨
t Regensburg, Germany
Physarum polycephalum is a versatile organism, dis-
playing several alternative cell types and developmental
transitions [1]. Uninucleate amoebae and multinucleate
plasmodia constitute the two vegetative growth phases
in the life cycle. These two cell types differ in cellular
organization, behaviour and gene expression. In
adverse conditions, amoebae reversibly transform into
cysts. Usually, when the conditions are favourable,
amoebae mate and develop into plasmodia. Plasmodia
survive adverse conditions by transforming into
another kind of cysts, spherules. When starved in the
light, sporangia are formed. In favourable conditions,
spores hatch to release amoebae, thus completing the
cycle.
Of the various cell types in the life cycle of
P. polycephalum, only the plasmodium contains the
water soluble polymer, b-poly(l-malate) (PMLA)
[2–4]. The polymer is concentrated in the nuclei in
an amount comparable with that of DNA and hi-
stones [5]. Due to its structural similarity to the
backbone of nucleic acids, PMLA has been proposed
to bind nuclear proteins and function in a molecular

of mycetozoa and recently as a biogenic matrix for the synthesis of devices
for drug delivery. The acellular slime mold Physarum polycephalum is charac-
terized by two distinctive growth phases: uninucleated amoebae and multi-
nucleated plasmodia. In adverse conditions, plasmodia reversibly transform
into spherules. Only plasmodia synthesize poly(malic acid) (PMLA) and
PMLA-hydrolase (polymalatase). We have performed suppression subtrac-
tive hybridization (SSH) of cDNA from amoebae and plasmodia to identify
plasmodium-specific genes involved in PMLA metabolism. We found cDNA
encoding a plasmodium-specific, spherulin 3a-like polypeptide, NKA48
(spherulin 3b), but no evidence for a PMLA-synthetase encoding transcript.
Inhibitory RNA (RNAi)-induced knockdown of NKA48-cDNA generated a
severe reduction in the level of PMLA suggesting that spherulin 3b func-
tioned in regulating the level of PMLA. Unexpectedly, cDNA of poly-
malatase was not SSH-selected, suggesting its presence also in amoebae.
Quantitative PCR then revealed low levels of mRNA in amoebae, high levels
in plasmodia, and also low levels in spherules, in agreement with the expres-
sion under transcriptional regulation in these cells.
Abbreviations
DSDM, diluted semidefined medium; PMLA, b-poly(
L-malate); RNAi, inhibitory RNA; SDM, semidefined medium; Sph, spherulin; SSH,
suppression subtractive hybridisation.
1046 FEBS Journal 273 (2006) 1046–1055 ª 2006 The Authors Journal compilation ª 2006 FEBS
degraded to l-malate by a plasmodium-specific hy-
drolase (polymalatase) [4,5,10,11].
PMLA is a highly interesting polymer: applications
in pharmacy and medicine are proposed ([6] and refer-
ences therein); nanoconjugates of PMLA can be used
as drug delivery vehicles [12], the crystal structure is
being investigated [13]. However, little is known about
the regulation of the polymer at the genetic level of its

tive hybridization (SSH) using cDNAs of plasmodial
extracts as tester and of amoebal extracts as driver. A
large number of transcripts were found, most of them
false-positive and only three true-positive. One had a
high degree of identity with spherulin 3a and appeared
to be involved in regulation of PMLA levels in vivo.
None of the SSH-generated DNAs showed similarity
with a sequence listed in the databases that would be
indicative of a PMLA synthetase. Quantitative PCR
revealed that polymalatase mRNA was expressed at
considerably lower levels in amoebae and spherules
than in plasmodia. This paralleled contents of poly-
malatase protein [4,10] suggesting regulated expression
at the transcriptional level.
Results
Isolation of differentially expressed cDNAs
After SSH, differentially expressed cDNAs were ana-
lysed after two rounds of PCR. The amplified products
from the secondary nested PCR were ligated with
pGEM
Ò
T-vector and were transformed into DH10B
competent cells. About 70 white colonies were
obtained in total, 52 of which were selected. Plasmid
DNAs were isolated and analysed after restriction
enzyme digestion. Each DNA sequence occurred only
once in agreement with the fact that 5¢-ends of
mRNAs had been isolated with the Capfinder oligo-
nucleotides. Restriction to 5¢-ends was thought to
reduce the complexity of bands after SSH and enhance

degree of identity on the levels of nucleotides (84%)
and amino acids (86%) with spherulin 3a (Figs 1 and
2), the most abundant encystment-specific protein [16],
and identities with sequences of bc-crystallins (Fig. 2).
The total number of amino acids is 103, correlating
with a calculated molecular mass of 11271.5 and a the-
oretical isoelectric point of 4.88. Because of the high
similarity, the polypeptide encoded by NKA48 was
named spherulin 3b.
Knockdown of mRNA to NKA48 (spherulin 3b)
Macroplasmodia were injected with dsRNA to
NKA48 (spherulin 3b) and harvested after 24 h. Two
negative controls were performed: macroplasmodia
N. Pinchai et al. Cell type expression of spherulin 3b and polymalatase
FEBS Journal 273 (2006) 1046–1055 ª 2006 The Authors Journal compilation ª 2006 FEBS 1047
without microinjection and macroplasmodia injected
with unspecific dsRNA (generated using part of the
pGEM
Ò
-5zf(+) vector as template [14]). The degree of
mRNA knockdown was analysed by real-time PCR
with actin mRNA as reference. Figure 3 shows that
the ratio of mRNA to NKA48 over that of actin was
significantly reduced to 1% (P<0.001). Control
microinjection with unspecific dsRNA showed no
effect on mRNA levels (P > 0.5), indicating that the
knockdown was specific. The fact that this low residual
level was obtained after 24 h suggested that the half-
life of spherulin 3b mRNA was in the range of one to
a few hours and much less than the half-life of 24–

not observed.
Level of polymalatase mRNA at different stages
in the life cycle
The amount of polymalatase transcript at different sta-
ges in the life cycle was monitored by real-time PCR
using specific primers. Since the mRNA level of house-
keeping genes, such as of actin, varies from one cell
type to the other, a cloned fragment of the polymala-
tase gene was used as an external standard and subjec-
ted to the same treatment as the samples. In Fig. 3C,
Fig. 1. Nucleotide sequence alignment of spherulin 3b (1) with spherulin 3a (2). Identical residues are highlighted in grey. Start and stop
codons are given in bold. Forward and reverse primer for RNAi experiments are underlined. Ac, Accession Number. The alignment was car-
ried out using
CLUSTALW from and BLAST from NCBI.
Cell type expression of spherulin 3b and polymalatase N. Pinchai et al.
1048 FEBS Journal 273 (2006) 1046–1055 ª 2006 The Authors Journal compilation ª 2006 FEBS
the level of cDNA of polymalatase (corresponding to
the level of mRNA) was very low for amoebae and
spherules in comparison with plasmodia (P<0.001).
The expression of the gene in amoebae and plasmodia
explained, why cDNA was absent after SSH screening
(see above). The presence of cDNA in the stages of the
life cycle indicated that the protein could have some
general function. High levels specifically in plasmodia
are consistent with a functional affiliation to PMLA
and with a regulation of gene expression at the tran-
scription level.
Discussion
Physarum polycephalum belongs to the mycetozoa, the
multicellular eukaryotes more closely related to ani-

spherulation of plasmodia and that then comprise
 10% of the total mRNA [16]. Among them, spheru-
Fig. 2. Structural alignment of amino acid sequence by motifs, of spherulin 3b with spherulin 3a and other members of the bc-crystallin fam-
ily: (1) spherulin 3b from P. polycephalum; (2) spherulin 3a from P. polycephalum; (3) hypothetical protein YPTB2846 from Yersinia pseudo-
tuberculosis; (4) hypothetical protein YmolA_01000341 from Y. mollaretii; (5) hypothetical protein Y1348 from Y. pestis; (6) c-crystallin from
Danio rerio; and (7) development-specific protein S homologue from Myxococcus xanthis. The residues highlighted in black indicate glycines,
serines, and aromatics that are conserved in a bc-crystallin fold. The residues shown in grey indicate the side chains and backbone sites that
are involved in calcium binding [24]. The residues in the conserved tyrosine corners are in bold [24]. ‘Greek key’ motifs are highlighted with
underlines: single line, first motif; broken line, second motif; double line, third motif. Motif searches were performed using
PROSITE from
. Similarity search and multiple alignment were carried out using
CLUSTALW from and BLAST
from NCBI. Ac, Accession Number.
N. Pinchai et al. Cell type expression of spherulin 3b and polymalatase
FEBS Journal 273 (2006) 1046–1055 ª 2006 The Authors Journal compilation ª 2006 FEBS 1049
lin 3a is the most abundant mRNA. During differenti-
ation, synthesis of PMLA discontinues, and the
remaining polymer is exported into the extracellular
fluid and degraded. It is assumed that the PMLA syn-
thesizing enzyme is downregulated at the onset of
spherulation.
Despite considerable effort, knowledge of PMLA
synthetase activity and its regulation is still fragment-
ary [15]. To discover stage-specifc genes affiliated with
PMLA metabolism and ultimately to identify the syn-
thetase gene, SSH was used with mRNA of the plas-
modium as the tester and mRNA of amoebae as the
driver. The amoebae strain chosen was LU352, which
was not identical with plasmodia of strain M
3

to address the evolutionary origin of the vertebrate
bc-crystallins [24]. Typically, two successive Greek
key motives (underlined in Fig. 2, each approximately
40 amino acid residues) pair to form a domain. The
domain fold contains a pair of calcium binding sites.
While the bc-crystallins of lens (not shown) and lower
organisms in Fig. 2 contain two domain folds, spheru-
lin 3a and NKA48 contain only a single domain fold.
The stability of these two proteins is highly dependent
on calcium binding [25]. The typical domain motives
contain a ‘tyrosine corner’ in the domain centre as
seen in proteins 3–6 of Fig. 2 or slightly displaced as
A
B
C
Fig. 3. Knockdown experiments and stage specific expression of
polymalatase mRNA. (A) Knockdown of NKA48 mRNA by specific
dsRNA. Levels of mRNA relative to that of actin are shown 24 h
after microinjection with dsRNA to NKA48 and with unspecific con-
trol dsRNA to pGEM-5zf(+) vector. Standard deviations refer to
experiments in triplicates. (B) PMLA content of plasmodia injected
with dsRNA to NKA48 in the RNAi experiment. The data are refer-
enced to protein contents. Standard deviations are shown for
measurements in triplicates. (C) mRNA levels of polymalatase in
different cell types during the life cycle. Levels were measured
in terms of cDNA by PCR referenced to a standard as described in
Experimental procedures. One-hundred per cent mRNA (plasmodia)
refers to 8.91 pgÆlL
)1
standard cDNA. Standard deviations are

this dsRNA treatment. However, because spherulin 3a
is not transcribed in the plasmodium [16], this possibil-
ity could not have effected the suppression of PMLA
synthesis.
Among other possibilities, this effect on PMLA
synthesis could be the result of loss of induction at
the transcriptional level, of loss of activation of the
synthetase protein itself, or of derepression of
enzyme(s) catalysing PMLA degradation. An interest-
ing interplay of NKA48 with spherulin 3a could be
imagined if both proteins bound competitively at the
same loci but only NKA48 was an inducer and ⁄ or
activator. In a physiologically meaningful mechanism,
spherulin 3a would displace NKA48 during the onset
of spherulation and suppress PMLA synthesizing
activity.
Degradation of PMLA during the onset of spherula-
tion is catalysed by enzymatically active forms of
polymalatase in the extraplasmodial fluid [5,10,11].
During plasmodia growth, only catalytic amounts of
polymalatase are contained in the culture medium,
while large amounts of zymogen reside within the plas-
modium. Correspondingly, zymogen and polymalatase
with different functions have been proposed, namely a
PMLA hydrolysing variant in the exterior and a chap-
eroning adapter variant in the interior of plasmodia
[7,10,11]. Polymalatase activity depends on zymogen
activation [10] at the outer surface of plasmodia
(unpublished results). The enzymology has been inves-
tigated in detail [5,10,11,26,27].

grown for further 24 h and then harvested for the analyses
of mRNA and PMLA content.
Culture conditions for spherule preparation
Spherules were induced by the transfer of 2-day-old micro-
plasmodia to a non-nutrient salt medium and were shaken
in the dark for 2 days at 24 °C as described [30]. After
replacement with fresh salt medium, spherules were incuba-
ted at 24 °C for 1 day and were harvested for real-time
PCR.
Culture conditions for the growth of amoebae
DSPB plates (diluted SDM with phosphate buffer [28])
were inoculated with 3 · 10
5
amoebal cysts of the apogamic
strain LU352 [31], 100 lL formalin-killed bacteria, and
100 lL Millipore water. The plates were incubated at 24 °C
N. Pinchai et al. Cell type expression of spherulin 3b and polymalatase
FEBS Journal 273 (2006) 1046–1055 ª 2006 The Authors Journal compilation ª 2006 FEBS 1051
for 48 h to allow excystment and were then transferred to
30 °C. After 4 days at 30 °C the plates became confluent
and were harvested for SSH and real-time PCR.
RNA isolation
To isolate total RNA, amoebae and macroplasmodia were
harvested from the agar plates and immediately frozen in
liquid nitrogen. RNA isolation was carried out by using the
QIAGEN RNeasy
Ò
Mini Kit (Qiagen, Hilden, Germany)
and a maximum of 100 mg frozen cells.
PolyA

Suppression subtractive hybridization
Differentially expressed cDNAs in plasmodia and amoebae
were identified following the SSH technique described by
Diatchenko et al. [32]. Plasmodial extract mRNA was used
as tester and amoebal extract mRNA as driver. Only
poly(A)
+
RNA was used for first-strand cDNA synthesis.
PCR reactions were optimized and performed in such a
way that syntheses remained in the exponential phase. Care
was taken that at least 25% of total cDNA was ligated
with adaptors on both ends. The success of SSH was tested
for an abundantly expressed housekeeping gene (actin
Ppa35 [33], accession number M21500), for a less abun-
dantly expressed gene lig1 [34], and for the known stage-
specific genes actin-fragmin kinase [35], fragmin A [36],
fragmin P [36], and polymalatase (accession no. AJ543320)
using primers to the published cDNA sequences. Also, the
efficiency of SSH was checked by comparing the number of
PCR cycles necessary to produce equal amounts of actin
cDNA in probes containing equal amounts of either sub-
stracted or unsubstracted DNA.
Subtracted PCR products were then ligated with pGEM
Ò
T-vectors (Promega, Mannheim, Germany) and were trans-
formed into DH10B competent cells. The plasmids were
isolated using Nucleospin
Ò
Plasmid Kit (Machery-Nagel,
Du

GAAATGTCCGTCCAACAAGGAG-3¢ (forward) and 5¢-
GCCTTCTAATACGACTCACTATAGGGACCACGATG
ATGGATGAAATG-3¢ (reverse). Both primers contained
T7-polymerase promoter at the 5¢ terminus and were cus-
tom-synthesized by MWG-Biotech. The resulting 294-bp
DNA spanned the nucleotides 51–310 of the gene including
the origin of transcription, and was used as template for
in vitro dsRNA synthesis as described by Donze and Picard
[37]. In the case of NKA8, the forward primer was 5¢-GAT
GCATAATACGACTCACTATAGGGAGTGCCTTGCAA
GGAGTATTG-3¢ and the reverse primer was 5¢-GCCTTC
TAATACGACTCACTATAGGGAGCTCGTAATAGCTT
TTGGAC-3¢, the resulting DNA spanning nucleotides
21–536 of the gene (accession number DQ017262). For con-
trol injections, nonspecific dsRNA was generated by the
same method using a PCR-derived fragment with 592 bp,
nucleotides 142–734 from the vector pGEM(R)-5zf(+)
(Technical Servics, Promega Corporation, Madison, WI,
USA). Each knockdown experiment was carried out with
Cell type expression of spherulin 3b and polymalatase N. Pinchai et al.
1052 FEBS Journal 273 (2006) 1046–1055 ª 2006 The Authors Journal compilation ª 2006 FEBS
10 lg dsRNA, microinjected into the veins of macro-
plasmodia. After 24 h, the plasmodia were analysed by
real-time PCR.
Real-time PCR
The amount of NKA48-specific mRNA in the RNA inter-
ference experiment was measured with reference to mRNA
expressed for actin Ppa35 gene using the Roche-LightCycler
(Roche, Mannheim, Germany). cDNA was synthesized
from 2 lg total RNA of each sample and reference, and

T-vector
was ligated with 16 ng purified PCR product and was
transformed into DH10B competent cells (Bethesda
Research Laboratories, Frederick, MD). Plasmid isolation
was carried out using the Nucleospin
Ò
Plasmid Kit
(Machery-Nagel). For insert isolation, 12 lg plasmid DNA
was digested with NcoI and SpeIat37°C for 1.5 h and
was analysed on a 2% agarose gel. The purified DNA
fragment was used as standard.
Total RNA was isolated from cells of the different stages
in the life cycle, and first-strand cDNA synthesis was per-
formed in triplicate as above for mRNAs, but using 2 lg
total RNA. Real-time PCR was carried out with 40 ng
cDNA of each sample in parallel with five different
amounts of the standard DNA, using the same primer pair
for polymalatase as above and cycling conditions 95 °C for
15 min, followed by 35 cycles of 95 °C for 15 s, 58 °C for
20 s, and 72 °C for 20 s. Default settings of the Lightcycler
Software Version 3.5.3 and conditions in the linear range of
the PCR-reaction were ensured.
Quantitative analysis of PMLA
Macroplasmodia were harvested, weighed and transferred
into a glass homogenizer. Two vols lysis buffer (50 mm
Tris ⁄ HCl pH 7.5, 5 mm NaS
2
O
5
,50mm EGTA, 10 mm

B-S & Holler E (2002) The DNA-polymerase inhibiting
activity of poly (b-L-malic acid) in nuclear extract dur-
ing the cell cycle of Physarum polycephalum. Eur J
Biochem 269, 1253–1258.
4 Windisch C, Miller S, Reisner H, Angerer B, Achham-
mer G & Holler E (1992) Production and degradation of
b-poly-L-malate in cultures of Physarum polycephalum.
Cell Biol Internat Reports 16, 1211–1213.
5 Schmidt A, Windisch C & Holler E (1996) Nuclear
accumulation and homeostasis of the unusual polymer
b-poly (L-malate) in plasmodia of Physarum
polycephalum. Eur J Cell Bio 70, 373–380.
6 Lee B-S, Vert M & Holler E (2002) Water-soluble ali-
phatic polyesters: Poly(malic acid)s. In Biopolymers
Volume 3a Polyesters (Doi, Y & Steinbu
¨
chel, A, eds),
pp. 76–103. Wiley-VCH, Weinheim, Germany.
7 Karl M, Gasselmaier B, Krieg RC & Holler E (2003)
Localization of fluorescence-labeled poly(malic acid) to
N. Pinchai et al. Cell type expression of spherulin 3b and polymalatase
FEBS Journal 273 (2006) 1046–1055 ª 2006 The Authors Journal compilation ª 2006 FEBS 1053
the nuclei of the plasmodium of Physarum polycepha-
lum. Eur J Biochem 270, 1536–1542.
8 Karl M, Anderson R & Holler E (2004) Injection of
poly(b-L-malate) into the plasmodium of Physarum
polycephalum shortens the cell cycle and increases
growth rate. Eur J Biochem 271, 3805–3811.
9 Lee B-S & Holler E (1999) Effects of culture conditions
on beta-poly(L-malate) production by Physarum polyce-

277.
17 Baldauf SL, Roger AJ, Wenk-Siefert I & Doolittle WF
(2004) A kingdom-level phylogeny of eukaryotes based
on combined protein data. Science 290, 972–977.
18 Baldauf SL & Doolittle WF (1997) Origin and evolution
of the slime molds (Mycetozoa). Proc Natl Acad Sci
USA 94, 12007–12012.
19 Angerer B & Holler E (2005) Large complexes of
b-poly(L-malate) with DNA polymerase a, histones,
and other proteins in nuclei of growing plasmodia
of Physarum polycephalum. Biochemistry 34, 14741–
14751.
20 Doerhoefer S, Khodyreva S, Safranov IV, Wlassoff
WA, Anarbaev R, Lavrik OI & Holler E (1998) Mole-
cular constituents of the replication apparatus in the
plasmodium of Physarum polycephalum: identification
by photoaffinity labelling. Microbiology 144, 3181–3193.
21 Chet I & Rush HP (1969) Induction of spherule forma-
tion in Physarum polycephalum by polyols. J Bacteriol
100, 674–678.
22 Wistow G, Summers L & Blundell T (2005) Myxococcus
xanthus spore coat protein S may have a similar struc-
ture to vertebrate lens beta-gamma-crystallins. Nature
315, 771–773.
23 de Jong WW, Caspers GJ & Leunissen JAM (1998)
Genealogy of the a crystallin-small heat shock protein
superfamily. Int J Biol Macromol 22, 151–162.
24 Clout NJ, Kretschmar M, Jaenicker R & Slingsby C
(2001) Crystal structure of the calcium-loaded spherulin
3a dimer sheds light on the evolution of the eye lens

Gurskaya N, Sverdlov ED et al. (1996) Suppression
subtractive hybridization: a method for generating
differentially regulated or tissue-specific cDNA probes
and libraries. Proc Natl Acad Sci USA 93, 6025–
6030.
33 Hamelin M, Adam L, Lemieux G & Pallotta D (1988)
Expression of the three unlinked isocoding actin genes
of Physarum polycephalum. DNA 7, 317–328.
34 Kroneder R, Cashmore AR & Marwan W (1999)
Phytochrome-induced expression of lig1, a homologue
of the fission yeast cell-cycle checkpoint gene hus1,
is associated with the developemental switch in
Cell type expression of spherulin 3b and polymalatase N. Pinchai et al.
1054 FEBS Journal 273 (2006) 1046–1055 ª 2006 The Authors Journal compilation ª 2006 FEBS
Physarum polycephalum plasmodium. Curr Genet 36,
86–93.
35 Gettemans J, de Ville Y, Vandekerckhove J & Waelkens
E (1993) Purification and partial amino acid sequence
of actin-fragmin kinase from Physarum polycephalum.
Eur J Biochem 214, 111–119.
36 T’Jampens D, Bailey J, Cook L, Constantin B, Van-
deckerckhove J & Gettemans J (1999) Physarum
amoebae express a distinct fragmin-like actin-binding
protein that controls in vivo phosphorylation of actin by
the actin-fragmin kinase. Eur J Biochem 265, 240–250.
37 Donze O & Picard D (2002) RNA interference in mam-
malian cells using siRNAs synthesized with T7 RNA
polymerase. Nucleic Acids Res 46, 1–4.
38 Bradford M (1976) A rapid and sensitive method for
the quantitation of microgram quantities of protein util-