Simultaneous, bidirectional translocation of 32P and 33P between wood blocks connected by mycelial cords of Hypholoma fasciculare


  • Björn Lindahl,

    Corresponding author
    1. Department of Forest Mycology and Pathology, SLU, Box 7026, SE-750 07 Uppsala, Sweden;
      Author for correspondence: Björn Lindahl Tel: +46 18 67 27 25 Fax: +46 18 30 92 45
    Search for more papers by this author
  • Roger Finlay,

    1. Department of Forest Mycology and Pathology, SLU, Box 7026, SE-750 07 Uppsala, Sweden;
    Search for more papers by this author
  • Stefan Olsson

    1. Department of Ecology, Section of Genetics and Microbiology, RVAU, Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark
    Search for more papers by this author

Author for correspondence: Björn Lindahl Tel: +46 18 67 27 25 Fax: +46 18 30 92 45


  • • Translocation of nutrients through fungal mycelia has mainly been viewed in terms of unidirectional movement from sources to sinks. This paper describes an experiment designed to test the possibility of bidirectional phosphorus transport in mycelial cords of the white rot fungus Hypholoma fasciculare growing in soil microcosms.
  • • A novel technique is described that uses the different energy spectra of the radioactive phosphorus isotopes 32 P and 33 P, in combination with nondestructive electronic autoradiography, to monitor the distribution of the two isotopes within single mycelial cord systems. The tracer isotopes were applied to pairs of wood blocks connected by mycelial cords, and isotope translocations were monitored during a 29-d period.
  • • Significant amounts of phosphorus were transported in both directions simultaneously between the two colonized wood blocks.
  • • Simultaneous, bidirectional translocation of phosphorus occurs in mycelial cords of H. fasciculare . Such circulation of nutrients in fungal mycelia should equalize differences in nutrient concentrations in different parts of the mycelia, enabling net translocation of nutrients from areas of high nutrient availability to those of high nutrient demand.


Although fungi are commonly included in the diverse group of life forms called microorganisms and are often studied by microbiologists, many fungal mycelia are large. Both saprotrophic (Thompson & Rayner, 1983) and ectomycorrhizal (Finlay & Read, 1986) mycelia can extend over distances on a decimetre to metre scale. These large mycelia are generally differentiated into diffuse mycelium and more or less organized mycelial cords. Through experiments using radioactive tracer isotopes, mycelial cords have been shown to be translocative pathways for nutrients and carbohydrates (Jennings, 1987; Cairney, 1992; Boddy, 1999). The manner in which nutrients and carbohydrates are transported in cords has been subject to much discussion.

Translocation through mycelial cords would enable resources to move from areas of excess –‘sources’ to areas of demand –‘sinks’. Unidirectional transport of nutrients from sources to sinks has been suggested to follow bulk flow of water in the cords (Jennings, 1987). Brownlee & Jennings (1982) studied acropetal transport of 14C added as glucose or aspartic acid, 32P added as phosphate and 42K in cords of Serpula lacrymans extending from wood blocks over a nutritionally inert substrate. When the osmotic potential of the solution surrounding the wood blocks was lowered, transport stopped. This led the authors to suggest that pressure driven bulk flow of water was the mechanism of nutrient transport.

Results from several other experiments, however, suggest that phosphorus in fungal mycelia can move both acropetally and basipetally (Wells et al., 1990; Olsson & Gray, 1998; Wells et al., 1998a; Lindahl et al., 1999). Granlund et al. (1985) found simultaneous bidirectional transport of carbohydrates in single rhizomorphs of Armillaria mellea using 3H and 14C labelled glucose. If nutrients can move through mycelial cords in opposite directions simultaneously, they would be able to circulate throughout the mycelial cord system, as speculated by Wells et al. (1998b) and Olsson (1999). This view of mycelial translocation implies that transport of nutrients does not occur unidirectionally as a response to source – sink polarities in the mycelium. Instead, transport occurs in all possible directions independently of polarities in the mycelium, but concentration gradients cause flux differences in different directions that result in net translocation from sources to sinks. At sites of uptake, nutrients or carbohydrates would be added to the mobile pool, increasing the concentration of the substance in question and causing a net transport away from that site. At sites of high nutrient demand, the transported substances would be converted to nonmobile forms, decreasing the concentrations of the transported forms and causing a net transport towards that site.

Harley & Smith (1983 ) suggest that cytoplasmic streaming may be an important mechanism of translocation in mycorrhizal fungi. Cycling of the cytoplasm within the fungal cells would enable bidirectional translocation of substances. Cairney (1992 ) suggests a model in which apoplastic, pressure-driven bulk flow of water and dissolved nutrients occurs in large diameter ‘vessel hyphae’ at the same time as symplastic translocation, driven by cytoplasmic streaming and concentration gradients, occurs in other hyphae within the same cord. Using fluorescent probes, Shepherd et al. (1993a ) showed the presence of a system of tubular vacuoles in Pisolithus tinctorius . These highly motile structures seem to be a universal feature of fungi ( Rees et al., 1994 ) and have been seen to transport their content with peristaltic-like movements in two directions simultaneously within the same hyphae. The tubular vacuoles also penetrate the septal pores, facilitating cell to cell transport ( Shepherd et al., 1993b ). However, the system of tubular vacuoles has so far only been demonstrated within a few mm of the hyphal apices. Bidirectional transport could also be carried out in vesicles moving on tubulin filaments ( Steinberg, 1998 ).

In the present study, electronic autoradiography was used to demonstrate simultaneous bidirectional translocation of phosphorus added as 33P- and 32P-labelled orthophosphate to wood blocks connected by mycelial cords of the white rot fungus Hypholoma fasciculare. Electronic autoradiography is a useful method to study translocation of radiotracers in intact fungal mycelia (Timonen et al., 1996; Olsson & Gray, 1998; Lindahl et al., 1999). A new technique was developed in which the different isotope radiation energy spectra enabled nondestructive measurements of the distribution of the two isotopes within mycelial cord systems in soil microcosms.

Materials and Methods

Experimental design

Microcosm systems were set up in 20 × 2 × 1 cm trays made of transparent plastic. In each tray, a 1-mm-thick strip of Perspex was covered with sieved (2-mm-mesh) forest soil from the organic horizon, which was slightly compressed to a thickness of approx. 3 m. This arrangement made it possible to lift the perspex strip and soil out of the tray for autoradiographic scanning (Lindahl et al., 1999). The soil was collected from a mixed coniferous forest outside Uppsala, Sweden with a field layer mainly composed of Pleurozium schreberi (Brid.) Mitt. and had a pH of 4. Living mosses and freshly fallen litter were removed before collection. The soil was stored at +4°C until used. Each microcosm was wetted with 2 ml of deionized water.

Wood blocks (diameter 15 mm, thickness c. 2 mm) were prepared from a freshly felled stem of Betula pendula Roth. and stored at −18°C until used. The wood blocks were autoclaved twice (125°C, 30 min) with a 2-d interval. Agar plugs with growing mycelium of H. fasciculare (Fr.) Kumm. (strain LH16 isolated in Lund, Sweden) were placed in the centre of Petri dishes with Hagem agar (Stenlid, 1985). After a mycelium had developed on the agar, wood blocks were added in front of the growing mycelial edges. The Petri dishes were incubated at 25°C for 11 d until the wood blocks were covered with mycelium.

At one end of each microcosm system, some soil was removed, and an inoculated wood block was placed so that the upper surface of the block was level with the soil surface. The microcosms were incubated in darkness at 17°C. Water loss through desiccation was compensated for gravimetrically each week by addition of deionized water with a Pasteur pipette. After 9 wk of incubation, new wood blocks (baits) were placed immediately in front of the growing mycelia that had extended c. 12 cm from the inoculated blocks. The systems were incubated for another 13 d until the fungi had fully colonized the baits, which were then connected to the inoculated blocks by mycelial cords.

The microcosms were randomly divided into two groups with 6 replicates in each. In the control microcosms, the mycelial cords that connected the wood blocks were cut halfway between the wood blocks, and a 5-mm broad strip of soil was removed to prevent mycelial contact and nutrient translocation between the two wood blocks. The microcosms of the test group were left with intact connective mycelial cords.

Tracer isotopes were supplied to the mycelium covering the wood blocks as 10 µl droplets of deionized water containing 0.1 MBq 32P-orthophosphate (carrier free) in the case of the baits and 0.4 MBq 33P-orthophosphate (110TBq mmol−1) in the case of the inoculated wood blocks. After 24 h the droplets, which still remained on the hydrophobic mycelial surface, were sucked up with a piece of tissue paper.


Radioactivity in the wood blocks was measured nondestructively at 8-d intervals using a Packard Instant Imager electronic autoradiography system (Packard Instrument Co, Meriden, CT, USA) (Lindahl et al., 1999). The contribution of each of the two isotopes to the total measured radioactivity in the wood blocks was estimated by scanning each microcosm system twice, the second time with the microcosm covered with 5 sheets of acetate overhead transparency film (0.1 mm sheet−1) in addition to the 6-µm-thick mylar film (Packard re-usable sample covers 6000597) used to protect the detector. This drastically changed the counting efficiency for the two isotopes. For each wood block and each scanning occasion a transmittance factor T was calculated:

image(Eqn 1)

(AT, the measured activity with 5 sheets of transparency film covering the microcosm; and AO, the measured activity without the plastic sheets.) The control systems were used to find T-values for wood blocks containing only one of the two isotopes.

If A32 and A33 are the respective contributions of 32P and 33P to the activity measured without plastic sheets, and T32 and T33 are the transmittance factors of wood blocks containing only one of the isotopes, then the following calculations can be used to estimate the contribution of each individual isotope to the total radiation.

image(Eqn 2)


image(Eqn 3)

Substitution gives:

image(Eqn 4)


image(Eqn 5)

The formulae above were also used to prepare images of the distribution of individual isotopes in the microcosms. After ‘smoothing’ the two original autoradiographic images of a microcosm to decrease the random variation between pixels (each pixel was calculated as the average of 81 surrounding pixels in the original images), the formulae above were applied to each pixel, making it possible to calculate the contribution of each isotope to the cpm value for each pixel in the autoradiographs.


After 29 d, the wood blocks were removed from the systems, cut into small pieces and digested in a mixture of 2.5 ml 70% perchloric acid and 2.5 ml 68% nitric acid for 60 min at 65°C. From each of the obtained solutions, two 250 µl subsamples were taken and transferred to glass scintillation vials. The subsamples were then bleached with 250 µl of 35% hydrogen peroxide for 30 min at 50°C. Scintillation cocktail (18 ml Hionic flour, Packard Instrument Co, Meriden, CT, USA) was added and the activities of the two isotopes were measured in a Packard TriCarb 2100 TR scintillation counter using Full Spectrum dpm analysis (Packard Instrument Co, Meriden, CT, USA). All activity estimates and measurements were corrected to take account of decay and the differences in half-life of the two isotopes.


After addition to the mycelium, the 32P tracer was not evenly distributed throughout the wood blocks, but remained immobilized in the mycelium at the point of addition. Even at the end of the experiment, activity was concentrated to the area beneath the feeding droplet (Fig. 1).

Figure 1.

Changes in the distribution of radioactivity within a wood block colonized by Hypholoma fasciculare after 32 P was added to the mycelium in a small water droplet placed centrally on the wood block. After an uptake period of 24 h the droplet was removed. The radioactivity was immobilized at the centre of the wood block, spreading only slowly towards the periphery. The figures in the legend represent time after addition of the radioactive isotope. The approximate diameter of the droplet is indicated in the figure by a bar. The wood block was a bait from a control system with severed cords.

Translocation of tracer isotopes between the wood blocks could be visualized with the help of electronic autoradiography. In the inoculated blocks of systems with intact cords, the transmittance factor T increased from 0.02 at 5 d after labelling to 0.10 at the end of the experiment, indicating import of high energy radiating 32P into the 33P labelled inoculated blocks. In the control systems with severed cords the transmittance factor of the inoculated blocks remained at 0.005. In the baits of systems with intact cords the transmittance factor decreased from 0.54 at 5 d after labelling to 0.50 at the end of the experiment, indicating import of low energy radiating 33P into the 32P labelled baits. In the control systems the transmittance factor of the baits increased from 0.54 to 0.55 (Fig. 2).

Figure 2.

Changes of transmittance factor (proportion of activity transmitted through 5 sheets of plastic film) of radiation from wood blocks colonized by Hypholoma fasciculare in soil microcosms and supplied with 32 P- or 33 P-labelled phosphate. (a) Baits supplied with 32 P. Open circles represent wood blocks in microcosm systems where mycelial cords connect the baits to inoculated wood blocks, supplied with 33 P. Closed circles represent systems where connecting mycelial cords have been cut. (b) Inoculated wood blocks supplied with 33 P. Open circles represent wood blocks in microcosm systems where mycelial cords connect the inoculated blocks to baits, supplied with 32 P. Closed circles represent systems where connecting mycelial cords have been cut. Error bars show SE and asterisks indicate significant differences between the control systems (cut mycelial cords) and intact systems.

With T33 set to 0.005 and T32 set to 0.55, the contributions of the two isotopes to the total activity could be estimated (Fig. 3). The activity of 33P in the baits increased throughout the experiment, and after 29 d 21% of the total 33P activity in the wood blocks was estimated to have been translocated from the inoculated blocks to the baits. The activity of 32P in the inoculated blocks increased to an estimated 7% of the total 32P activity in the wood blocks 29 d after tracer addition (Fig. 4). These figures were confirmed by scintillation counting, which showed that the electronic autoradiography method slightly underestimated the translocated fractions of both isotopes. At harvest, 24% of the total 33P in the wood blocks had been translocated from the inoculated blocks to the baits and 9% of the total 32P in the wood blocks had been translocated from the baits to the inoculated blocks. In the control systems, the two isotopes were quite well confined to the wood blocks to which they were added. At harvest an average of 0.3% (max. 0.8%) of the isotopes was found in the wood block to which they were not added (except for one system which was removed from analysis due to substantial cross-contamination during harvest).

Figure 3.

Autoradiographic images of the distribution of radioactivity in two wood blocks interconnected by cords of Hypholoma fasciculare in a soil microcosm. The lower wood block – the inoculated block – was supplied with 33 P-phosphate and the upper wood block – the bait – was supplied with 32 P–phosphate. The images were acquired 29 d after addition of the tracer isotopes. The left image (a) shows the distribution of 32 P, which is translocated from the addition point at the bait to the inoculated block and to the growing mycelial front at the top of the image. The right image (b) shows the distribution of 33 P, which is translocated from the addition point at the inoculated block to the bait and further to the growing mycelial front.

Figure 4.

Electronic autoradiographic estimates of changes in radioactivity in wood blocks interconnected by mycelial cords of Hypholoma fasciculare in soil microcosms. One of the wood blocks – the inoculated block – was supplied with 33 P-phosphate and the other wood block – the bait – was supplied with 32 P-phosphate. Closed circles represent the fraction of the total 33 P activity translocated from the inoculated blocks to the baits and open circles represent the fraction of the total 32 P activity translocated from the baits to the inoculated blocks. Error bars show SE.


Phosphorus was transported in both directions simultaneously between two colonized wood blocks connected by mycelial cords of H. fasciculare. This result is consistent with an earlier study by Wells et al. (1998a) using separate baits for Phanerochaete velutina labelled with a single isotope, although this finding was not discussed by the authors. It also supports earlier observations from mycelia growing on agar (Olsson & Gray, 1998). In this respect, phosphorus behaves in the same way as glucose, which is transported bidirectionally in rhizomorphs of Armillaria (Granlund et al., 1985).

Bidirectional transport would enable circulation of transported substances within the mycelium (Olsson, 1999). At sites of uptake, substances are added to the circulation stream. Substances are immobilized from the circulation stream in chemical forms that are not transported, but may also be reconverted from immobile forms back to transported forms. Circulation of substances could be facilitated by peristaltic movements of tubular vacuoles (Shepherd et al., 1993a) or with the help of vesicles moving bidirectionally on tubulin filaments (Steinberg, 1998). Unidirectional transport within individual hyphae but in opposite directions in different hyphal elements would also enable circulation within the mycelium. In a system like this, absolute transport of a compound could be much larger than the net translocation (Olsson, 1999). If transport in any direction is nonselective, that is, vacuoles or vesicles are loaded with a representative sample of the mobile pool at the loading site, then net translocation of nutrients would occur along concentration gradients from sources to sinks.

The immobilization of added tracer in the mycelium at the point of addition is in agreement with the results of Olsson & Gray (1998), who found that 32P added to agar cultures was immobilized at the point of addition. Clipson et al. (1987) also found that only a small fraction of 32P, added to rhizomorphs in a field experiment, was translocated away from the feeding site. This indicates that immobilization of phosphorus into storage vesicles or stationary compounds, for example precipitated polyphosphate (Bücking & Heyser, 1999), nucleic acids or phospholipids, is a rapid process. Although the tracer isotopes were rapidly immobilized, there was significant translocation between the wood blocks during most of the experiment (Fig. 4), indicating conversion of immobilized phosphorus back to mobile forms which were transported throughout the mycelium. These mobile forms could be orthophosphate, short length polyphosphate (Ashford et al., 1994) or free nucleotides.

No conclusions can be drawn from this experiment about the net translocation of nonradioactive phosphorus or the source–sink relationship between the two wood blocks, since the sizes of the total phosphorus pools in the wood blocks are unknown.

The idea that nutrients circulate bidirectionally in fungal mycelia is supported by experiments in which mesh bags with 15N-labelled pine needles were incubated in the field or in litter mesocosm in the lab (Berg, 1988). Absolute amounts of nitrogen increased in the litterbags but the amounts of 15N decreased. The tracer isotope was translocated away from the litterbags, although the bags were sinks for nitrogen. On the forest floor, basidiomycetous mycelia interconnect various resource units with large differences in substrate quality. Rapid circulation of nutrients throughout the mycelium and associated net transport of nutrients from sources to sinks would enable fungi to integrate over the spatial heterogeneity of their environment. Rapid mycelial growth, for example in newly colonized litter units, would decrease cytoplasmic nutrient concentrations, as nutrients are incorporated into structural tissues. Thus, there would be a net translocation of nutrients to rapidly growing mycelium. In mycelium colonizing litter components in late stages of decomposition with small amounts of easily accessible carbohydrates remaining, the incorporation of nutrients into structural tissues would decline and concentrations of mobile nutrients in the cytoplasm would increase, leading to a net translocation away from these components. This theory agrees with long-term field studies of nutrient dynamics in decomposing needles (Staaf & Berg, 1982). Traditional models of nutrient cycling often use the nutrient concentration of single litter units to predict the potential for nutrient immobilization or release. If nutrients can move rapidly throughout the mycelia of decomposer fungi, models of nutrient circulation must consider the nutritional status of the whole mycelium instead of single resource units.


Financial support from the Swedish Council for Forestry & Agriculture Research (SJFR), the Swedish Natural Sciences Research Council (NFR) and the Knut and Alice Wallenberg Foundation is gratefully acknowledged.