Antonie van Leeuwenhoek 61: 221-229, 1992. �9 1992 Kluwer Academic Publishers. Printed in the Netherlands.

Adhesive knob formation by conidia of the nematophagous fungus Drechmeria coniospora

P.H.J.F. van den Boogert, 1 J. Dijksterhuis, 2 H. Velvis l & M. Veenhuis 2 l DLO-Institute for Soil Fertility Research, P.O. Box 30003, 9750 RA Haren, The Netherlands 2 Laboratory for Electron Microscopy, University of Groningen, Kerklaan 30, 9751 NN Haren, The Netherlands

Received 18 June 1991; accepted 5 November 1991

Key words: adhesive knobs, conidiogenesis, Drechmeria coniospora, nematodes, nematophagous fungi

Abstract

We studied conidiogenesis and adhesive knob formation (maturation) by newly developed conidia of the nematophagous fungus Drechmeria coniospora. Upon conidiogenesis on infected nematodes or during saprophytic growth of the fungus in axenic cultures compact clusters of conidia developed. Less than 10% of such clustered conidia matured; mature conidia were invariably located on the periphery of the clusters.

The kinetics and rate of maturation of conidia were studied in in vitro systems and in soil. In both cases adhesive knobs were formed; the rate at which knobs were formed appeared to be determined by the age of the conidia, the temperature and the soil moisture. In addition, knob formation was suppressed at increasing conidial densities. Under favorable conditions, however, over 90% of the conidia matured within a period of 3 days. The rate of knob formation was neither influenced by the presence of nematodes nor by that of exogenous nutrients, which suggests that maturation is an autonomous process. Electron-microscopical analysis indicated that budding of the conidia at the initial stage of maturation occurred simultaneously with the deposition of the sticky, adhesive layer around the wall of the developing knob.

The ecological significance of the time- and spatially separated maturation of conidia after conidiogenesis is discussed with respect to survival of the conidia.

Introduction

Drechmeria coniospora (Drechsler) W. Gains & Jansson is a nematophagous fungus that infects various species of bacterivorous, insectivorous and plant-parasitic nematodes through adhesive coni- dia (Dtirschner 1983; Gams & Jansson 1985; Poinar & Jansson 1986; Hashem 1988). The life cycle of D. coniospora has been described in detail by Saikawa (1982), Jansson et al. (1984) and Dijksterhuis et al. (1990, 1991). Upon digestion of the nematode con- tents numerous conidia are produced outside the nematode body (Drechsler 1941; Barron 1977).

Remarkably, the newly developed conidia are not immediately infective since they do not have an adhesive knob at the time they are released from the conidiiferous peg. However, since capturing is strictly dependent on the presence of these knobs, a subsequent maturation of newly formed D. co- niospora conidia is an essential prerequisite for infection of following host nematodes.

Our current research programme was designed to study the possible use of D. coniospora for the biological control of plant-parasitic nematodes, so mass production of conidia in axenic culture (Loh- mann & Sikora 1990) was necessary. However,

222

little is known about the maturation process of conidia under natural and cultural conditions.

In this paper we present information on i) coni- diogenesis and conidial maturation by D. conio- spora using light and electron microscopy and ii) the influence of various biotic and abiotic factors on the maturation process. The ecological signif- icance of the observed mechanism of conidiogene- sis and conidial maturation in relation to survival of D. coniospora in the soil is discussed.

Materials and methods

Organisms and cultivation methods

Drechmeria coniospora, kindly provided by Dr. H.B. Jansson, University of Lund, Sweden, was used throughout this study. The following culti- vation media were used: malt extract peptone agar (MPA), containing 15.0g malt extract (Oxoid L39), 1.0 g peptone (Oxoid L72) and 12 g agar (Ox- oid L13) per i deionized water; liquid malt peptone (MP), see MPA without agar; corn meal agar (CMA), containing 2.0 g corn meal extract and 15 g agar (Oxoid CM103) per 1 deionized water; liquid corn meal (CM), see CMA without agar; water agar (WAP), containing 12.0 g purified agar (Oxoid L28) per 1 glass-distilled water; mineral solution (MS), containing 2.5 g NH4SO4, 0.2 g MgSO4"7H20, 3.0g NaH2PO4 and 0.7g K2HPO4 per 1 deionized water.

The bacteria-feeding nematode Panagrellus re- divivus (L.) Goodey was used for infection studies. The organism was axenically cultured, harvested and washed according to Jansson & Nordbring- Hertz (1979). The nematode suspension was trans- ferred to tap water before use.

Isolation of conidia

Drechmeria coniospora was cultured at 21~ on MPA or CMA in 90 mm diam Petri dishes, or in liquid cultures in 100 ml Erlenmeyer flasks supple- mented with 40 ml MP or CM. Conidia were col- lected from agar plates by streaking with a Drigal-

ski spatula in a few ml MS. In order to separate conidia from the mycelial fraction, the crude sus- pension was heavily agitated and subsequently fil- tered through a double nylon cloth with a pore size of 40 tzm. The concentration of the conidia in the suspension was determined using a hemacyto- meter.

Development of adhesive knobs

The process of knob formation was studied in standing cultures with MPA, CMA or WAP and in shaking cultures with MP or CM. In standing cul- tures the conidia were distributed over 45-mm diam polycarbonate membrane filters (0.22ram pore size) by vacuum filtration at - 80 kPA. The seeded filters were placed on 3 mm thick agar lay- ers in 60 mm diam Petri dishes. At regular time intervals the filters were removed from the Petri dishes, air-dried and cleared with paraffin oil for microscopic observations.

For shaking cultures 100ml Erlenmeyers were used on a rotary shaker (200rpm) supplemented with 40 ml liquid medium and desired concentra- tions of conidia. Samples were taken at regular time intervals.

Knob formation was furthermore studied in gno- tobiotic microcosms containing sandy soil and na- tive micro-organisms. Prior to use the soil was sieved through a 3-mm screen, air-dried to 1.6MPa, gamma-sterilized (6Mrad) and subse- quently reinoculated. For that purpose the inocu- lure was prepared from a 1 : 10 (w/w) soil suspen- sion in sterile tap water after passing through 1.2 mm pore diam filters by vacuum filtration at - 80 kPa. The water retention characteristics were determined in 5-cm diam cylinders filled with 100 g soil at a bulk density of 1.1Mg m -3 according to gravity and vacuum pressure methods. The amount of water at a certain soil water tension was derived from the retention curve. Subsequently aliquots of 20 g soil were adjusted to desired water tensions by adding sterile tap water to the reinfected soil. Poly- carbonate membrane filters (Sartorius, 45mm diam and 22 mm pore size) seeded with conidia were enveloped between two layers of this soil in

60-mm diam Petri dishes. The Petri dishes were sealed with parafilm to prevent water evaporation. At regular time intervals filters were removed and prepared for light microscopy as described above.

The influence of the presence of nematodes on knob formation was studied in two-compartment wells separated by a 0.4mm pore polycarbonate membrane (Costar; 35mm diam multi-transwell plates). The transwell system contained suspen- sions of conidia in the lower compartment and P. redivivus in the upper compartment.

Electron microscopy

Conidia were fixed in 1.5% KMnO4 for 20min at room temperature. After staining in 1% aqueous uranyl acetate the material was dehydrated and embedded in E 812. Ultrathin sections were cut on a LKB-Ultratome with a diamond knife and exam- ined in a Philips CM 10 electron microscope. For low-temperature scanning electron microscopy, nematodes at different stages of infection with D. coniospora were used and prepared as described in Dijksterhuis et al. (1991).

Data analys&

Each experiment was designed in triplicate; per object, fold and sample date 300 non-clustered co- nidia were evaluated for the presence of knobs. The ratios between the numbers of knob-bearing conidia and the total numbers, expressed as per- centages, were recorded. Percentages were statisti- cally analyzed after arcsin transformation. The nu- merical data were subjected to analysis of variance (ANOVA) using the Genstat 5 programs (copy- right 1987, Lawes Agricultural Trust, Rothamsted Experimental Station). Differences between means were evaluated using the least significant difference (LSD) at 5% confidence level.

223

Knob-bearing conidia (% of total number) 100 ]

8~ 1 6oi

2O

0 T

0 5 10 1~5

LSDI~ !- 2 35

�9 liquid �9 solid

2o 25

Time (days)

Fig. 1. Knob formation by D. coniospora conidia (knob-bearing conidia as percentage of total number; N = 300) in shaking and standing cultures with liquid and solid corn meal medium re- spectively at 21 ~ C. LSD indicates the least significant difference at P = 0.05 in three replications.

Results

Conidiogenesis

Conidiogenesis on infected P. redivivus resulted in the formation of compact clusters of conidia (up to 40 each) developed around the conidiiferous pegs (Jansson et al. 1984; Dijksterhuis et al. 1991). In vitro, on MPA or CMA, a similar clustering of conidia was observed. Light-microscopical obser- vations indicated that relatively few conidia, locat- ed on the periphery of the clusters, had developed knobs.

To study conidiogenesis in more detail, hyphal development, conidial production and adhesive knob formation were compared in liquid and on solid CMA, following inoculation with 105 conidia per ml medium. Under in vitro conditions the ma- jority of the conidia failed to germinate. On aver- age, less than 0.1% of the introduced conidia devel- oped into fungal micro-colonies. Germ tube devel- opment occurred preferentially at the apical region of the conidium as illustrated in Fig. 3A. These micro-colonies produced numerous, mostly imma- ture, knobless conidia, except at the start of the growing period when a relatively large proportion of the newly produced conidia developed knobs

224

Fig. 2. Low-temperature scanning electron micrographs of initial stages of conidiophore formation and clustering in D. coniospora. (A, B) primordia of developing adhesive knobs (indicated by arrows). (C) maturing (arrows) liberated conidia, located on the nematode surface. (D) large clusters of conidia; knob-bearing conidia, located on the periphery of the clusters, are evident. The marker represents 5/zm.

(Fig. 1). The conidial yield, expressed as the num- ber of spores per g nutrient powder added, calcu- lated from stationary phase cultures at day 25, amounted to 9.9 and 9.0 log units on solid and in liquid CM, respectively. On solid CM, however,

knob formation remained low compared with the

liquid medium (Fig. 1).

225

Fig. 3. Thin sections of KMnO4-fixed conidia of D. coniospora. (A) germination of a mature conidium in vitro as outgrowth of the adhesive knob (arrow). (B-E) different stages of adhesive knob development. (B) survey showing the initial development of the adhesive layer (arrow); details of subsequent stages in C and D. Note the additional electron-dense layer in the wall of the knob (arrow). (E, F) fully developed knobs; the radiated structure of the adhesive is clearly visualized and especially evident at negatively stained intact conidia. (Abbreviations: L - lipid droplet; N - nucleus. The marker represents 0.5 ~m).

226

Knob-beadn 9 conidia (% of total number) 100

80

60-

40.

20.

0

F~I PA

I CMA

0

Knob-bearing conidia (% of total number) 100

LSDcs%)= 5.30 LS[~s%)=2.35

80

60 Il l 10 days

1 [~ 31 days 4O

B 1 2 3 5 6 7 13

Time (days) Log conidia (ml'l)

Fig. 4. Knob formation by D. coniospora conidia (knob-bearing conidia as percentage of total number; N = 300) on corn meal agar (CMA) and purified water agar (PA) at 21 ~ C. LSD indicates the least significant difference at P = 0.05 in three replications.

Microscopical observations

In a separate study we have shown that D. co- niospora produces high numbers (up to 5.104 coni- dia per single nematode) of conidia after infection of the nematode P. redivivus (Dijksterhuis et al. 1991). Conidia were produced on conidiiferous pegs. Each peg produced one conidium within 4 h. The majority of the newly produced conidia re- mained adhered to each other (Fig. 2A), resulting in the formation of large clusters (Fig. 2D). Prior to their separation from the conidiiferous peg knob initiation was never observed. Knob formation started early after the conidia were liberated from the conidiiferous peg: on pegs bearing chains of three or more conidia, the firstly formed conidium showed signs of knob initiation (Fig. 2A). Also, early formed conidia liberated from the peg as observed on the nematode surface frequently showed knob initiation (Fig. 2C). Our results sug- gested that the rate of knob formation remained restricted and predominantly took place on the initial conidia formed and on those which were located on the outside of the developing cluster (Fig. 2D). These observations were confirmed by analysis of series of ultrathin serial sections through large clusters present at the final stage of nematode digestion: knob-bearing conidia were solely pre-

Fig. 5. Knob-bearing D. coniospora conidia (knob-bearing coni- dia as percentage of total number; N = 300) at different conidial densities and at different age, determined in shake cultures with mineral solution after 5 days of incubation at 21 ~ C. LSD indicates the least significant difference at P = 0.05 in three replications.

sent on the periphery of such clusters. Similar ob- servations with respect to the occurrence and distri- bution of mature conidia were made on clusters of artificially grown conidia (e.g. on CMA). The morphogenesis of the adhesive knob was studied in detail after incubation of CMA-grown conidia in distilled water. Light microscopical studies indicat- ed that maturation of individual conidia generally took 8-10 h under these conditions. Observations on ultrathin sections of KMnO4-fixed cells revealed that budding of the conidium occurred simultane- ously with the development of the adhesive layer (Fig. 3B). The walls of the buds were characterized by the presence of an additional central thin elec- tron-dense layer invariably absent in the normal cell wall (Fig. 3C). During further development both the size of the bud and the amount of the adhesive material deposited on its outside in- creased (Fig. 3C, D); after maturation the adhesive layer measured approximately 0.5/zm in width (Fig. 3E) and showed a typical radiated structure, as was especially evident after negative staining of intact conidia (Fig. 3F).

Knob.bearing conidia (% of total number)

1 O0

80

60

40

20

Colony diam.

LSD(s%~= 6.71

5 10 15 20 25

ram) - - 20

0 30

Temperature (o C)

15

10

5

Fig. 6. Colony diameters (ram) and conidial knob formation

(knob-bearing conidia as percentage of total number ; N = 300)

in D. coniospora at different temperatures , determined on malt

pepton agar after 25 days and on water agar after 2 days,

respectively. LSD indicates the least significant difference at

P = 0.05 in three replications.

Factors affecting the process of knob formation

The effect of exogenic nutrients on conidial knob formation was studied on WAP and CMA, using conidia collected from 4-week-old agar cultures, at a density of approximately 500 conidia per mm 2 agar surface. The first knobs appeared within 24 hours after the conidia had been introduced to the new environments; subsequently their number gradually increased in time (Fig. 4). After 3 days of incubation over 80% of the conidia had developed knobs indicating that maturation could occur effec- tively in the absence of exogenic nutrients.

The possible effects of nematodes and/or nema- tode excretion products on knob formation were studied in two-compartment transweIls containing 1 ml spore suspension (1.5 105 conidia per ml) in the lower compartment and 1 ml (1000 individuals) P. redivivus suspension in the upper compartment. We observed no significant differences in the rate of knob formation, whether nematodes were pre- sent in the system or not; after two days of in- cubation at 21~ the number of mature conidia amounted to 81% as compared to 79% in the con- trols.

The effect of inoculum age and density on knob

227

100

80

60

40

20

0

Knob-bearing conidia (% of total number) LSD(5%}= 3.89~

i !-

4.7 3.3 2.6 117

i /

i Day 7

, Day 2 /Day 1

Log water tension (kPa]

Fig. 7. Knob formation of D. coniospora conidia (knob-bearing

conidia as percentage of total number ; N = 300) in soil at various water tensions (log kPa) at 21 ~ C. LSD indicates the least

significant difference at P = 0.05 in three replications.

formation was determined in shake flask cultures on MS. The results indicated that knob formation decreased significantly at increasing densities of the inoculum (Fig. 5). This density effect was de- pendent on the physiological state of the inoculum conidia, since it became less pronounced when old- er conidia were used (Fig. 5). Similar density ef- fects were found in standing cultures on WAP (not shown). To determine whether soluble conidial ex- udates were associated with the density effect, fresh conidia were incubated (final concentration 100 per ml) in liquid MS (which previously con- tained l0 s conidia per ml) for two weeks. The re- sults indicated that the rate of knob formation in this medium did not differ significantly from that in the fresh MS. Apparently, knob formation was reversibly suppressed at increasing numbers of co- nidia.

The effect of temperature on hyphal growth was determined on MPA using 2-week-old micro-colo- nies on agar discs as inoculum; the effect of temper- ature on knob formation was determined on WAP inoculated at 1500 conidia per mm 2. The growth response to temperature showed an optimum be- tween 20 and 25 ~ C (Fig. 6). Adhesive knob forma- tion showed a temperature response similar to that of radial growth (Fig. 6).

The effect of soil moisture on knob formation

228

was determined in soil with different water ten- sions at 21~ inoculated with 3200 conidia per mm2-filter membrane. As depicted in Fig. 7, the rate of knob formation gradually increased up to almost 100% at decreasing water tensions. The maximum water tension at which knob formation occurred amounted to 3.5-4.7 log units kPa.

Discussion

We have studied the influence of different biotic and abiotic factors on adhesive knob formation by newly formed conidia of D. coniospora. As sug- gested before (Jansson et al. 1984; Dijksterhuis et al. 1991), knob formation, i.e. conidial maturation, is spatially separated from conidiogenesis and oc- curs temporarily after the newly formed conidium is released from the conidiiferous peg. Our present results strongly suggested that knob formation is dependent on local conidial densities. This was indicated by the findings that during conidiogenesis at either the expense of digested nematodes or during saprophytic growth of the fungus in axenic cultures, where high local densities (clusters) are achieved, less than 10% of newly formed conidia produced adhesive knobs. This density effect was also observed upon incubation of artificially grown conidia in liquid cultures where knob formation was reversibly suppressed at increasing numbers of conidia. However, in shaking cultures under condi- tions, in which clustering of conidia was effectively prevented, knob formation readily proceeded up to 80%. Knob formation was not influenced by either exogenous nutrients or the presence of host nematodes. These facts support the assumption that conidial maturation is primarily an autono- mous process, controlled by ecophysiological par- ameters (temperature and soil moisture). Under optimal conditions with respect to conidial age and density, artificially grown conidia achieve 80-100% maturity within 3 days upon incubation both in vitro as well as in soil. However, conidia are revers- ibly impeded to mature when they aggregate into clusters as achieved during conidiogenesis at the expense of digested nematodes or during sapro- phytic growth. Whether it is physical or chemical

signals that regulate maturation remains to be elu- cidated.

As D. coniospora is almost completely dependent on host nematodes in nature (Barron 1977) and adhesive conidia are the only infective entities, the question arises why the fungus does not solely pro- duce adhesive conidia immediately upon infection of nematodes. In earlier work (Dijksterhuis et al. 1990, 1991) we showed that the main strategy of D. coniospora during infection of nematodes includes the production of conidia rather than mycelial mass. The separation of adhesive knob formation from conidiogenesis is in line with this view, in that it substantially reduces the time required for the generation of single conidia. Furthermore, the ob- served suppression of the maturation step may be advantageous to the fungus in terms of long-term survival. Upon dispersion of conidia from the clus- ters, for instance due to mechanical force (such as movement of soil-inhabiting animals or water), new conidia will mature and become infective. The relatively rapid initial development of adhesive knobs ensures infection of following host nema- todes. Hence, we speculate on the significance of conidial clustering as a mode to regulate matura- tion, which thus benefits survival of D. coniospora in nature.

Acknowledgements

The first and second author are supported by the Netherlands Integrated Soil Research Programme (project No. C8-8b).

We gratefully acknowledge the kind hospitality of Dr. R.A. Samson and M.I. van der Horst (Cen- traalbureau voor Schimmelcultures, Baarn, the Netherlands) and Drs. P. Staugaard for the expert advice and help during low-temperature scanning electron microscopy studies.

References

Barron GL, (1977) The nematode destroying fungi. Topics in

Mycobiology No. 1, Canadian Biological Publications Ltd, Guelph, Ontario, Canada, 140 pp

Dijksterhuis J, Veenhuis M & Harder W (1990) Ultrastructural study of adhesion and initial stages of infection of nematodes by conidia of Drechmeria coniospora. Mycolog. Research 94: 1-8

Dijksterbuis J, Harder W, Wyss U & Veenhuis M (1991) Colo- nization and digestion of nematodes by the endoparasitic nematophagous fungus Drechmeria coniospora. Mycolog. Research 95:873-878

Drechsler C (1941) Some hypbomycetes parasitic on free-living terricolous nematodes. Phytopathol. 31:773-802

D/irsschner U (1983) Pilzliche Endoparasiten an beweglichen Nematodenstadien. Mitteilungen aus der biologischen Bun- desanstalt ffir Land- und Forstwirtschaft, Berlin-Dahlem 217, 83 pp

Gains W & Jansson H-B (1985) The nematode parasite Meria coniospora Drechsler in pure culture and its classification. Mycotaxon 22:33-38

Hashem MH (1988) Untersuchungen zur Wirtsspezifit~it und zum Entwicklungszyklus endoparasit~irer Nematodenpilze an

229

beweglichen Nematoden. Dissertation, Christian Albrechts University, Kiel, Germany

Jansson H-B, Hofsten A von& Mecklenburg C vom (1984) Life cycle of the endoparasitic nematophagous fungus Meria co- niospora: a light and electron microscopic study. Antonie van Leeuwenhoek 50:321-327

Jansson H-B & Nordbring-Hertz B (1979) Attraction of nema- todes to living mycelium of nematopbagous fungi. J. Gen. Microbiol. 112:89-93

Lohmann U & Sikora RA (1989) Mass production of the endo- parasitic fungi Drechmeria coniospora, Verticillium bala- noides and Harposporium anguillulae in liquid culture. Ne- matologica 35:97-104

Poinar Jr GO & Jansson H-B (1986) Infection of Neoaplectana and Heterorhabditis (Rhabdita : Nematoda) with the preda- tory fungi, Monacrosporium ellipsosporum and A rthrobotrys oligospora (Moniliales : Deuteromycetes). Rev. N6matol. 9: 241-244

Saikawa M (1982) An electron microscope study of Meria co- niospora, an endozoic nematophagous Hyphomycete. Can. J. Bot. 60:2019-2023