with type 1 anatomy, had quite negative osmotic potentials in their ST-CC complex when compared with
mesophyll cells, thereby demonstrating that all of the previously listed criteria for phloem loading must
be retained.
Van Bel [284] and Turgeon [285] proposed a mechanism by which plasmodesmata could act as an
osmotic trap. In that scheme, sucrose and galactinol, the precursors of raffinose and stachyose, would dif-
fuse from mesophyll cells into intermediary or companion cells, where the larger oligosaccharides would
be synthesized. Resultant oligosaccharides would be too large to return to the mesophyll cells through
plasmodesmata but could penetrate the larger diameter plasmodesmata into sieve tubes. This would pro-
vide the osmotic gradient reported for cucurbits and Coleus. Such a system is supported by the distribu-
tion of raffinose and stachyose in various leaf cells of melon [286]. However, as pointed out by Haritatos
et al. [286], failure of melon to translocate galactinol is not explained by this model, nor is the fate of myo-
inositol, the product of raffinose and stachyose synthesis. It may be that there is an efficient retrieval
(apoplastic?) system for these materials. One must also wonder how a higher concentration of hexose and
a lower concentration of sucrose are maintained in mesophyll cells than in the ST-CC [286] if sucrose can
diffuse through connecting plasmodesmata unless the tonoplast is the diffusion-limiting structure.
The mechanism proposed by Van Bel [284] and Turgeon [285] would require that much of the car-
bon translocated by plants with type 1 anatomy would be as large oligosaccharides. Correlation between
species that transport various sugars (Ref. 34 and others) and the vascular bundle types [269] is quite
strong, although there are striking exceptions. For example, Fraxinus ornus, the most extreme of the type
1 group, transports not only the higher oligosaccharides but also significant amounts of mannitol. In ad-
dition, both grape and willow have type 1 anatomy yet translocate sucrose as the primary carbohydrate
[134,148]. This mechanism of phloem loading suggested by Van Bel [287] also does not explain the ob-
servation that the ratio of various amino acids in sieve tube exudate does not match that of mesophyll cells
[159]. This lack of correspondence, however, could be accounted for by transfer of amino acids from
xylem to phloem [257] using an apoplastic pathway.
D. Summary, Phloem Loading
Briefly, the apoplastic model of phloem loading requires that assimilates within mesophyll cells move
symplastically from cell to cell via plasmodesmata until they reach minor vein parenchyma cells. Assim-
ilates are then transferred to the apoplast, where they are absorbed by the ST-CC complex using the H-
sucrose symport and driven by the proton gradient maintained by the H-ATPase. It is seen as essential
that movement be symplastic up to the vein, for water moving from the vein to the point of transpiration
in the apoplast would move assimilate away from the phloem. This system requires type 2 vein anatomy.
The mechanism for the symplastic pathway is not so clear. This system requires type 1 vein anatomy.
The pathway is through plasmodesmata from mesophyll cells into sieve tubes. A mechanism of size trap-
ping for selection of raffinose family oligosacchrides has been proposed. However, this does not explain
the mechanism for selection against other materials. It would be attractive to suggest that there are pro-
teins that transport selected materials through plasmodesmata as reported for nucleic acids [164,166];
however, there is no datum suggesting such a mechanism. In addition, data on the diffusion of dyes must
be explained, although these may be an artifact [276].
Much work is need before we understand the processes of the plasmodesmata. Current studies in var-
ious laboratories using molecular procedures offer the greatest promise for solving problems of symplas-
tic phloem loading as well as other processes involving cell-to-cell communication. This new informa-
tion will be useful in understanding both mechanisms of phloem loading, for even the apoplastic pathway
appears to involve transport through plasmodesmata up to the ST-CC.
We are left with the conclusion that both symplastic and apoplastic phloem loading do occur but that
the mechanism of only the apoplastic pathway is well understood (see Ref. 287), for McLean et al. [288]
in their review stated that “our understanding of plasmodesmatal structure and function is still naive.”
VI. PHLOEM UNLOADING
Assimilates are unloaded from phloem of the plant producing the assimilate either into vegetative cells
via the apoplast or plasmodesmata or into a developing embryo and/or endosperm (i.e., a separate plant)
[289–293]. Patrick [293] has classified these as apoplastic, symplastic, and maternal/filial. In special sit-
438 HENDRIX