Previous investigations have shown that axolotl double anterior thighs regenerate only a tapered extension of the femur after simple amputation, but undergo intercalary regeneration of the femur and a symmetrical tibia when a normal wrist blastema is grafted to the thigh. There are two possibilities to account for these results: (1) distal transformation in both terminal and intercalary regeneration depends upon special properties of posterior cells (such as polarizing ability) that are lacking in the anterior half of the limb, but which would be provided by the posterior half of a normal wrist blastema graft, or (2) the pattern of cellular interactions required for distal transformation during terminal regeneration is different from that required for distal transformation during intercalary regeneration.

These alternatives were tested by grafting axolotl double anterior wrist blastemas to double anterior thighs. The host thighs regenerated a femur and symmetrical tibia that structurally were the same as the bones intercalated after grafting a normal wrist blastema to a double anterior thigh. In all cases, the graft developed as a symmetrical hand comprising one to three carpals and one to two digits. These results rule out the necessity of any special properties of posterior cells for distal transformation or polarization of the anterior-posterior axis and suggest instead that distal transformation during intercalary regeneration involves a different pattern of cellular interactions than in terminal regeneration.

The rule of distal transformation is a formalization of the observation that cells of the amphibian limb regeneration blastema normally form only those structures distal to the level of amputation (Rose, 1962), even when regeneration takes place from a limb whose proximal-distal (PD) axis has been reversed (Butler, 1955). The cellular basis of this rule is that a continuous sequence of PD positional information (Wolpert, 1969, 1971) is recorded in the limb cells during ontogeny (Pescitelli & Stocum, 1981), and this cellular memory prevents blastema cells from forming morphological patterns proximal to their level of origin (Pescitelli & Stocum, 1980; Maden, 1980). This constraint ensures that the structural patterns formed during regeneration are always those which complete the limb distally.

Symmetrical morphological patterns are inimical in varying degrees to distal transformation (Bryant, 1976; Bryant & Baca, 1978; Stocum, 1978; Tank, 1978; Tank & Holder, 1978; Holder, Tank & Bryant, 1980). The most inhibitory pattern is a double anterior half (dA) upper arm or thigh. When these limbs are amputated 15–30 days after construction, they most often regenerate only a short cone of cartilage capped with soft tissue. In contrast, double posterior half (dP) thigh stumps amputated after the same healing interval are often able to regenerate proximodistally complete double posterior limbs which in the majority of cases are hypomorphic with regard to midline structures. Both dA and dP lower arms and shanks regenerate proximodistally complete limbs with loss of midline structures (Stocum, 1978; Krasner & Bryant, 1980).

The results of double half-limb regeneration have led to the formulation of a model in which distal transformation leading to restoration of the PD axis depends upon interactions between cellular positional values that specify pattern in the transverse plane (Bryant, 1978; Bryant & Baca, 1978; Holder, et al. 1980; Krasner & Bryant, 1980). In this model, which is a modification of the polar coordinate model of French, Bryant & Bryant (1976), the transverse limb pattern is represented by a series of numbers around the circumference of the limb and the PD axial pattern is represented by a series of letters along the radii of the circle formed by the circumference. Non-adjacent cells on the circumference (dermal connective tissue cells) are assumed to interact across short arcs to form a new complete circle by intercalary regeneration. The new circle, which lies inside the circumference, then adopts the next most distal PD (radial) value. Repetitions of this process restore all the missing PD values in a proximal to distal sequence.

The difference in the degree of distal transformation between dA and dP limbs is hypothesized to depend upon the extent of short-arc interactions between cells with the same or adjacent positional values on opposite sides of the midline. The pattern of short-arc interactions in dA limbs causes dedifferentiated cells having the same or adjacent positional values to quickly confront each other across the midline, eliminating all but a few of the most anterior values, halting intercalation and, therefore, distal transformation. In contrast, short-arc interactions in dP limbs are postulated to take place in a pattern that does not lead to such midline confrontations, thereby allowing intercalation to continue until all the PD levels of the limb are restored (Holder et al. 1980; Krasner & Bryant, 1980).

Restoration of missing PD positional values also takes place in normal limbs when distal blastemas are grafted to a more proximal stump level (Iten & Bryant, 1975; Stocum, 1975). The intercalated structures are derived from blastema cells contributed by the host stump (Pescitelli & Stocum, 1980; Maden, 1980). Furthermore, dA thigh stumps will regenerate missing intermediate structures when normal wrist blastemas are grafted to them (Stocum, 1980a). The latter result suggests that the distal transformation taking place after this operation is not dependent upon interactions across short arcs in the transverse plane, but is the result of direct intercalation of missing intermediate PD positional values between the values represented by the amputation levels of graft and host. Intercalary regeneration between two PD levels and terminal regeneration after simple amputation may therefore depend on different patterns of cellular interactions.

However, there is another hypothesis which might account for the difference in degree of terminal regeneration between dA and dP limbs and for the fact that PD intercalary regeneration from dA thighs is evoked by normal wrist blastemas. Distal transformation may require some special activity that is the exclusive property of cells in the posterior half of the limb. Thus, amputated dP limbs and dA limb stumps provided with a normal wrist blastema would undergo distal transformation due to the presence of posterior cells. Other evidence suggests that a special zone of posterior cells (zone of polarizing activity) is involved in polarization of the anterior-posterior axis of the embryonic urodele and amniote limb, and a dorsal zone may also be involved in polarizing the DV axis(Swett, 1938; Slack, 1976,1977,1980a; Fallon & Crosby, 1977; Summerbell & Tickle, 1977).

The latter hypothesis can be tested by experiments in which dA wrist blastemas are grafted proximal to their level of origin to either normal or dA thigh stumps. The results of such experiments are the subject of this paper.

White and dark axolotls (Ambystoma mexicanum), four months post-hatching and 60–70 mm in snout-tail tip length were used for all the experiments. In the first experiment, dA lower arms were surgically constructed on the right forelimbs of dark and white animals by replacing the posterior half of the lower arm with the anterior half of the left lower arm. The operations were done in petri dishes with the animals lying on paper towelling saturated with a 0·025% (w/v) solution of Chlorotone (Eastman Chemical Co.) made up in 100% Holtfreter solution, after first anaesthetizing the animals in a 0·075% (w/v) solution of Chlorotone made up in 1% Holtfreter solution. The donor and host limb halves were sufficiently adherent to one another after 2–4 h to transfer the animals to individual finger bowls containing 1% Holtfreter solution, where they were fed every other day with freshly-hatched brine shrimp. The dA lower arms were allowed to heal for 10 days, and were then amputated through the doubled carpus or distal zeugopodium. The right (normal) hindlimb was amputated through the midthigh at the same time.

In the second experiment, dA lower arms were first constructed on the right forelimbs of dark and white animals. After a 10-day healing period, dA thighs were constructed on the right hindlimbs of the same animals by grafting the skin and muscle from the anterior half of the left thigh in place of the skin and muscle of the posterior half of the right thigh. At the end of a further 10-day healing period, the dA forelimbs were amputated through the wrist and the dA hindlimbs through the midthigh.

In both experiments, when medium bud to early redifferentiation blastemas appeared (staging according to Stocum, 1979), the dA wrist blastemas were exchanged as autografts or homografts with either the normal or dA thigh blastemas, as diagrammed in Fig. 1.

Fig. 1

Diagram illustrating the exchange between dA wrist blastemas and normal or dA thigh blastemas. Hatched areas represent the anterior (A) half of the left limb that was grafted in place of the ppsterior (P) half of the right limb.

Fig. 1

Diagram illustrating the exchange between dA wrist blastemas and normal or dA thigh blastemas. Hatched areas represent the anterior (A) half of the left limb that was grafted in place of the ppsterior (P) half of the right limb.

In a number of cases in both experiments, dA donor wrists did not receive a reciprocal thigh graft, but were allowed to regenerate a second time so that the structure of their regenerates could be compared to that formed by dA wrist blastemas grafted to the thigh.

All limbs were allowed to regenerate for 40–50 days, after which they were removed from the animal, fixed in Gregg’s solution and stained for cartilage with methylene blue according to the Van Wijhe technique as modified by Gregg and Butler (Hamburger, 1960).

These experiments allowed an assessment of (1) the ability of normal and dA thigh blastemas to develop according to their origin on dA wrist stumps, and (2) the ability of dA wrist blastemas to evoke intercalary regeneration from normal or dA thigh stumps.

(1) dA wrist re-regenerates

Double anterior wrist stumps regenerating for the second time formed hands with one or two symmetrical digits and one to three symmetrically arranged carpals (Figs. 2, 3), and thus regenerated no differently from the grafted dA blastemas formed after the first amputation (see below).

Fig. 2

Donor dA wrist stump regenerated for the second time. Digit 1 is regenerated twice in a symmetrical arrangement. Three carpals, probably representing the radiale and carpal DI-2 (the most anterior carpals), regenerated. Note the loss of midline structures (middle column of carpals and digit 2) distal to the doubled radius (R).

Fig. 2

Donor dA wrist stump regenerated for the second time. Digit 1 is regenerated twice in a symmetrical arrangement. Three carpals, probably representing the radiale and carpal DI-2 (the most anterior carpals), regenerated. Note the loss of midline structures (middle column of carpals and digit 2) distal to the doubled radius (R).

Fig. 3

Another re-regenerate of a donor dA wrist showing extensive pattern convergence distal to the doubled radius (R). There are four carpal elements (arrow) arranged in a symmetrical pattern, and a single digit, probably composed of the anterior halves of two digit l’s fused in the midline.

Fig. 3

Another re-regenerate of a donor dA wrist showing extensive pattern convergence distal to the doubled radius (R). There are four carpal elements (arrow) arranged in a symmetrical pattern, and a single digit, probably composed of the anterior halves of two digit l’s fused in the midline.

(2) Normal and dA thigh blastemas grafted to dA wrist stumps

Three of five normal thigh blastemas developed according to origin when grafted to dA wrist stumps, forming a normal hindlimb distal to the wrist (Fig. 4). In one of the two remaining cases, the graft underwent resorption of prospective proximal structures, but redifferentiated a normal foot distal to the host zeugopodium. In the other case, the graft resorbed totally, and the host dA stump regenerated a short, tapered rod of cartilage.

Fig. 4

Normal thigh blastema autografted to a dA distal zeugopodium. A normal five-toed hindlimb developed distal to the doubled radius (R) of the host. H, Host humerus; F, T, f, femur, tibia and fibula of donor.

Fig. 4

Normal thigh blastema autografted to a dA distal zeugopodium. A normal five-toed hindlimb developed distal to the doubled radius (R) of the host. H, Host humerus; F, T, f, femur, tibia and fibula of donor.

The majority (7 of 12) of dA thigh blastemas grafted to dA wrist stumps also developed according to origin, redifferentiating a tapered rod of cartilage distal to the host wrist (Fig. 5). Of the remaining five cases, the graft failed to differentiate in two cases, resorbed in one case, followed by regeneration of a symmetrical single-digit hand from the host stump, and in two cases was pushed to one side as the host regenerated a symmetrical single-digit hand.

Fig. 5

dA thigh blastema homografted to dA distal zeugopodial stump (white animal to dark animal). The skin has been removed from the dorsal side of the specimen for better viewing of the skeleton. The blastema developed according to origin, forming a tapered cone of cartilage (arrow). The pigmentation pattern of the limb is coincident with the different morphological patterns of donor and host.

Fig. 5

dA thigh blastema homografted to dA distal zeugopodial stump (white animal to dark animal). The skin has been removed from the dorsal side of the specimen for better viewing of the skeleton. The blastema developed according to origin, forming a tapered cone of cartilage (arrow). The pigmentation pattern of the limb is coincident with the different morphological patterns of donor and host.

(3) dA wrist blastemas grafted to normal or dA thigh stumps

Table 1 summarizes the results of grafting dA wrist blastemas to normal thigh stumps. The grafts did not develop well. In half the cases, the graft resorbed and the host stump regenerated a normal hindlimb. In 20% of the cases, the host regenerated a normal hindlimb while the graft developed separately on the anterodorsal side of the host foot. In 30% of the cases, the graft formed the anterior digits of a host-donor chimaera (Fig. 6). There were no cases in which typical intercalary regeneration of the host stump took place so that normal hindlimb skeletal structures were intercalated up to the distal end of the tibiafibula, with the limb terminating in a symmetrical one- to two-digit hand.

Table 1

Results of grafting double anterior (dA) wrist blastemas to normal thigh stumps

Results of grafting double anterior (dA) wrist blastemas to normal thigh stumps
Results of grafting double anterior (dA) wrist blastemas to normal thigh stumps
Fig. 6

dA wrist blastema homografted to normal thigh stump (white animal to dark animal). The skin has been partially removed from the dorsal side of the specimen for better viewing of the skeleton. The host regenerated a femur, a short and somewhat abnormal tibia and fibula, a tarsal pattern exhibiting some abnormalities on the anterior side, and four toes numbered from anterior to posterior. The graft (arrow) formed two symmetrically arranged digit 1’s fused in the metacarpal region; these digits were in their proper location on the anterior side of the host foot and completed the digital array.

Fig. 6

dA wrist blastema homografted to normal thigh stump (white animal to dark animal). The skin has been partially removed from the dorsal side of the specimen for better viewing of the skeleton. The host regenerated a femur, a short and somewhat abnormal tibia and fibula, a tarsal pattern exhibiting some abnormalities on the anterior side, and four toes numbered from anterior to posterior. The graft (arrow) formed two symmetrically arranged digit 1’s fused in the metacarpal region; these digits were in their proper location on the anterior side of the host foot and completed the digital array.

Table 2 summarizes the results of grafting dA wrist blastemas to dA thigh stumps. Only 2 of 13 cases (both homografts) failed to exhibit typical intercalary regeneration of stylopodial and zeugopodial skeletal elements. In one of these two cases (white donor, dark host), the graft formed a single digit-like spike of unpigmented tissue on the non-regenerating host. At the time of fixation, the spike exhibited haemostasis, indicating that the donor tissue was undergoing chronic immunorejection. In the other case, the stylopodium and a single zeugopodial bone regenerated, but the limb did not terminate in carpals or digits, suggesting that while intercalary regeneration was proceeding, graft resorption was also taking place.

Table 2

Results of grafting double anterior (dA) wrist blastemas to double anterior thigh stumps

Results of grafting double anterior (dA) wrist blastemas to double anterior thigh stumps
Results of grafting double anterior (dA) wrist blastemas to double anterior thigh stumps

The remaining 11 cases (3 autograft, 8 homograft) formed regenerates consisting of a distal stylopodium (femur), a single symmetrical zeugopodial bone (tibia) and a one- to two-digit hand identical in structure to the hands regenerated by donor wrist stumps. Figs. 79 illustrate the skeletal structure of the three autograft cases.

Fig. 7

dA wrist blastema autografted to dA thigh stump. The host thigh intercalated the femur (F) and a single symmetrical tibia (T). The graft developed according to origin, forming a single carpal (C) and digit (D). The skin was removed from the specimen for better viewing of the skeleton.

Fig. 7

dA wrist blastema autografted to dA thigh stump. The host thigh intercalated the femur (F) and a single symmetrical tibia (T). The graft developed according to origin, forming a single carpal (C) and digit (D). The skin was removed from the specimen for better viewing of the skeleton.

Fig. 8

dA wrist blastema autografted to dA thigh stump. The thigh stump intercalated the femur (F), and a single symmetrical tibia (T). The graft formed one rodshaped and one round carpal (arrows) and a single digit. The skin was removed from the specimen for better viewing of the skeleton.

Fig. 8

dA wrist blastema autografted to dA thigh stump. The thigh stump intercalated the femur (F), and a single symmetrical tibia (T). The graft formed one rodshaped and one round carpal (arrows) and a single digit. The skin was removed from the specimen for better viewing of the skeleton.

Fig. 9

dA wrist blastema autografted to dA thigh stump. Again, the host intercalated femur (F) and tibia (T), while the graft developed a double anterior hand consisting of a single carpal (C) and two digit l’s fused along the longitudinal axis in the regions of the metacarpal and the first phalange. The skin was removed from the specimen for better viewing of the skeleton.

Fig. 9

dA wrist blastema autografted to dA thigh stump. Again, the host intercalated femur (F) and tibia (T), while the graft developed a double anterior hand consisting of a single carpal (C) and two digit l’s fused along the longitudinal axis in the regions of the metacarpal and the first phalange. The skin was removed from the specimen for better viewing of the skeleton.

The homograft cases, in general, did not develop as well as the autograft cases. Of the eight homografts, four displayed pigmentation patterns in which the donor colour was restricted to the hand and host colour to the intercalary regenerate (Figs. 10, 11, 12). This colour separation confirms the host origin of the intercalary regenerate. In two of these cases, the intercalated skeleton consists of a single bone which cannot be identified as either stylopodial or zeugopodial because it does not articulate with the femur (Figs. 11, 12).

Fig. 10

dA wrist blastema homografted to dA thigh stump (white animal to dark animal). The graft did not develop well, forming a single digit (arrow), but the host intercalated the femur (F) and a short tibia (T). Note that the host stump and intercalated region are covered with skin of host colour, while the digit has donor colour.

Fig. 10

dA wrist blastema homografted to dA thigh stump (white animal to dark animal). The graft did not develop well, forming a single digit (arrow), but the host intercalated the femur (F) and a short tibia (T). Note that the host stump and intercalated region are covered with skin of host colour, while the digit has donor colour.

Fig. 11

dA wrist blastema homografted to dA thigh stump (dark animal to white animal). The limb did not develop well. A single bone (arrow) was intercalated from the host stump (covered by skin of host colour), but whether this bone is femur or tibia cannot be determined with certainty. The graft developed a single digit with the donor pigmentation pattern, and exhibited intense signs of immunorejection (haemostasis) at the time of fixation.

Fig. 11

dA wrist blastema homografted to dA thigh stump (dark animal to white animal). The limb did not develop well. A single bone (arrow) was intercalated from the host stump (covered by skin of host colour), but whether this bone is femur or tibia cannot be determined with certainty. The graft developed a single digit with the donor pigmentation pattern, and exhibited intense signs of immunorejection (haemostasis) at the time of fixation.

Fig. 12

dA wrist blastema homografted to dA thigh stump (dark animal to white animal). The graft formed two digits fused along the longitudinal axis in the regions of the metacarpal and first phalange, and a single carpal element (arrow). The skin over these structures has the donor pigment pattern. A single bone (I) was intercalated from the host stump, but could not be identified as femur or tibia. The proximal two-thirds of this bone appeared to be embedded in the soft tissues of the host stump and its proximal one-third overlapped the distal end of the host femur stump (F). A small amount of fine pigment can be seen in the skin covering the distal one-third of the intercalated bone. The haemostasis of immunorejection was visible in the phalanges and distal half of the metacarpal. Blood vessels were dilated in the proximal half of the metacarpal region and in the carpal region, but blood flow was still excellent.

Fig. 12

dA wrist blastema homografted to dA thigh stump (dark animal to white animal). The graft formed two digits fused along the longitudinal axis in the regions of the metacarpal and first phalange, and a single carpal element (arrow). The skin over these structures has the donor pigment pattern. A single bone (I) was intercalated from the host stump, but could not be identified as femur or tibia. The proximal two-thirds of this bone appeared to be embedded in the soft tissues of the host stump and its proximal one-third overlapped the distal end of the host femur stump (F). A small amount of fine pigment can be seen in the skin covering the distal one-third of the intercalated bone. The haemostasis of immunorejection was visible in the phalanges and distal half of the metacarpal. Blood vessels were dilated in the proximal half of the metacarpal region and in the carpal region, but blood flow was still excellent.

The remaining four homograft cases (all dark donors and white hosts) exhibited donor pigmentation in the skin or around the blood vessels of the intercalary regenerate. The area occupied by donor melanocytes was usually restricted to either the distal end of the zeugopodium (one case) or to a thin band extending along a portion of the length of the zeugopodium (two cases, Fig. 13). In one case, however, pigment extended over the whole zeugopodium, the skeleton of which consisted of two radii fused in the longitudinal axis for most of their length. The pigmentation and skeletal patterns of the donor and host of this case are in accord, however, for the records show that the donor lower arm had been amputated through the doubled radius, at a more proximal level than other cases.

Fig. 13

Well-developed homograft of a dA wrist blastema to a dA thigh stump (dark animal to white animal). The graft formed two carpals in tandem (arrows) and a single digit. The host intercalated the femur (F) and a single symmetrical tibia (T). A thin band of pigmented skin lies on the dorsal side over the whole length of the zeugopodium. There was haemostasis in the phalangeal region, and the blood vessels in the metacarpal and carpal regions were heavily dilated, with sluggish blood flow. In the pigmented band of skin over the zeugopodium, the blood vessels appeared to be slightly dilated, but blood flow was normal.

Fig. 13

Well-developed homograft of a dA wrist blastema to a dA thigh stump (dark animal to white animal). The graft formed two carpals in tandem (arrows) and a single digit. The host intercalated the femur (F) and a single symmetrical tibia (T). A thin band of pigmented skin lies on the dorsal side over the whole length of the zeugopodium. There was haemostasis in the phalangeal region, and the blood vessels in the metacarpal and carpal regions were heavily dilated, with sluggish blood flow. In the pigmented band of skin over the zeugopodium, the blood vessels appeared to be slightly dilated, but blood flow was normal.

The fact that normal and dA thigh blastemas grafted to dA wrist stumps develop according to origin reinforces the conclusion, drawn from other work (Stocum, 1968,1980 b; Holder & Tank, 1979) that enough positional information to specify the pattern of the regenerate is inherited by the blastema from the pattern of the stump tissues during their dedifferentiation. The regenerate pattern is not primarily imposed upon or secondarily reinforced by a signal transmitted from the stump to the blastema after the latter has formed, as has recently been claimed by Slack (1980b).

Limbs composed of dA wrist blastemas grafted to normal thighs did not undergo intercalary regeneration. Instead, the host seemed to ignore the presence of the graft, or else the graft resorbed, with the host thigh regenerating a complete PD array of skeletal structures. In those cases where the graft survived, it was carried distally as the host thigh regenerated. The graft ultimately wound up at its normal position on the anterior side of the host foot, either integrated with the host toes to form a hand-foot chimaera, or as a separate structure on the anterodorsal side of the host foot. The positioning of the graft structures with respect to the host foot structures agrees with the finding of Shuraleff & Thornton (1967) that a foot blastema grafted to the thigh of a regenerating limb seeks its level of origin in the final regenerate. This fact strongly suggests that the molecular basis of positional information in the regenerating limb resides in the plasmalemma or surface coat of the limb cells. The reason for the high rate of resorption of the dA wrist blastema grafts is unknown. It is possible that they have difficulty in becoming established because their development is subordinated to and overwhelmed by the greater regenerative potential which may be afforded by the complete transverse positional value map at the amputation surface of the host.

In contrast to the above results, 85% of the limbs consisting of a dA wrist blastema grafted to a dA thigh stump exhibited typical intercalary regeneration, resulting in limbs consisting of a femur and/or a single, symmetrical tibia intercalated from the host stump, and terminating in a hand comprising a few symmetrically arranged carpals and one to two symmetrical digits. The hand structure was the same as that produced by dA wrist stumps that underwent reregeneration, and the symmetrical structure of the tibia is what would be expected from the fusion along the midline of two tibiae, each derived from one half of the double anterior-half thigh. The intercalary regenerates were exactly the same in structure as those observed after intercalary regeneration evoked by normal wrist blastemas grafted to dA thigh stumps (Stocum, 1980a).

Evidence that the graft redifferentiated according to its origin and for the host origin of the stylopodial and zeugopodial skeleton in these regenerates is provided by those homograft cases in which donor colour was restricted to the hand while the rest of the regenerate had host colour. In several cases, however, donor pigmentation was found over the distal end of the intercalary regenerate, or in a thin band extending along the longitudinal axis of the intercalary regenerate. The former pattern might reflect some mixing of graft and host cells in the most distal portion of the intercalary regenerate (see Pescitelli & Stocum, 1980). The latter pattern superficially might suggest that proximal transformation of graft cells took place. However, it is unlikely that this interpretation is valid. Studies of intercalary regeneration using ploidy markers have shown that the cells of the intercalated structures are derived solely from the host stump (Pescitelli & Stocum, 1980). The pigmented bands of skin can be explained by wound healing patterns in which fingers of donor epidermis bridge the grafthostjunction during initial wound healing. This epidermis could provide factors favourable for migration of melanocytes along it (see Frost & Malacinski, 1980, for review).

Since intercalary regeneration of dA thighs is evoked by grafts containing only a double anterior half-map of positional values, it is clear that cells in the posterior half of the urodele limb do not act in any special way to promote distal transformation, nor are they organized as a polarizing zone which specifies the anterior-posterior axial structure of the regenerate. Special polarizing or distal transformation-promoting properties of posterior limb cells are also ruled out by the finding that dP upper arms which have undergone prolonged healing times (30 days or more) after their construction exhibit regenerative behaviour characteristic of dA upper arms; they regenerate short, tapered spikes (Bryant, 1976; Tank, 1978; Tank & Holder, 1978). Carlson (1975) was also led to conclude that no polarizing zone exists in regenerating urodele limbs because he could observe no preferential location for the emergence of supernumerary limbs after skin rotation and amputation. We may also conclude that the inhibition of terminal regeneration in dA limbs is not due to a subthreshold innervation pattern in double anterior-half limbs. The fact that little or no pattern convergence occurs in the dA intercalary regenerates of the present experiments suggests instead that distal transformation during intercalary regeneration is accomplished by interactions occurring directly between PD positional values of graft and host cells without requiring any interactions between transverse values. However, it is not ruled out that transverse interactions may be taking place along with proximal-distal interactions during PD intercalary regeneration, but in a pattern that avoids convergence.

Either way, the results are compatible with the hypothesis of Bryant & Baca (1978) and Holder et al. (1980) that short-arc interactions take place between transverse positional values during terminal regeneration of the normal limb, and lead to pattern convergence in dA upper arms and thighs, and in dP upper arms that have healed for an extensive length of time. According to this hypothesis, the extent of distal transformation is proportional to the degree to which confrontation between like or adjacent transverse positional values across the midline is avoided by the pattern of short-arc interactions. Double anterior limbs healed for 15 days or more, and dP limbs healed for 30 days or more regenerate after amputation as if the pattern of short-arc interactions progressively eliminates midline positional values, leading to confrontations of transverse values inhibitory to further intercalation and halting distal transformation early in the course of blastemal outgrowth. In contrast, dP limbs healed for less than 30 days often regenerate as if the pattern of short-arc interaction does not eliminate as many midline values, allowing intercalation between transverse values to continue until all missing PD levels have been restored.

Presumably, there is a temporally related difference in healing patterns which makes it possible for dP limbs with ‘short’ healing times to undergo complete distal transformation, while dP limbs with ‘long’ healing times and dA limbs cannot. An effect of healing time has also been observed in dA forelimbs by Tank & Holder ( 1978), but is not as pronounced as that observed in dP forelimbs. When healed for less than 15 days after construction, dA upper arm stumps often regenerate past the elbow, but with longer healing times, the stumps either fail to regenerate at all, or regenerate a short spike of cartilage. Double anterior thighs healed for only 10 days always regenerate only a short cartilage spike (Stocum, 1978). In contrast to dP upper arm stumps, however, dA upper arm stumps never regenerate digits, regardless of healing time. Thus, Holder et al. (1980) postulate factors in addition to the pattern of healing to explain the difference in regenerative ability between dA and dP limbs amputated 5-15 days after construction. If it is assumed that the posterior half of the limb carries more positional values than the anterior half, confrontation of like or adjacent positional values across the midline will occur somewhat earlier during blastemal outgrowth in dA limbs than in dP limbs. Histological evidence consistent with this notion is that the posterior half of the limb contains more cells, at the same cell density, than the anterior half (Tank & Holder, 1980). Furthermore, Maden (1979) has found that posterior half-upper arms exhibit a greater regulative ability during terminal regeneration than anterior halves. At present, however, there is no histological or cytological evidence for the existence of interactions between cells across short arcs in the amputation plane. It would be of interest to look for such interactions using light microscopy, and scanning and transmission electron microscopy.

Although existing evidence is compatible with the idea that cellular interactions in the plane of amputation can lead to pattern convergence (deletion of midline structures) in double half-limbs, it does not justify the conclusion that convergent interactions actually lead to the cessation of distal transformation in these limbs; i.e., that the actual generation of PD positional values during regeneration is dependent upon interactions between transverse positional values. The convergent patterns in the regenerates of double half-limbs have been interpreted as being truly truncated, with the pattern ending in an abnormally proximal PD positional value. However, it is possible that what we have viewed as lack of distal transformation is an illusion created by convergence, the reality being that during terminal regeneration, positional values representing all the PD levels of the double half-limbs are reconstituted. However, due to progressive pattern convergence, only the most peripheral positional values are represented at the level of the zeugopodium and autopodium, giving a tapered, but actually proximodistally complete regenerate.

With this possibility in mind, we reasoned that if the latter notion has any validity, the distal tip of the blastema which forms on a dA thigh represents a PD positional value at least one and possibly two segments distal to the stylopodium. In a preliminary experiment, several dA thighs were constructed on large dark axolotls (80 – 90 mm snout-tail tip length). After a 10-day healing period, these limbs were amputated through the midthigh and a medium bud blastema allowed to form (approximately 12 days postamputation). The distal tip, comprizing about one-fourth the length of the blastema was then severed and grafted to the amputated normal thigh of a small white axolotl (25 – 30 mm snout-tail tip length). The exchange was made between large and small animals because when exchanges were made between animals of the same size, there was a wide disparity between cross-sectional areas of the graft and host, and the grafts always resorbed. Only one of three cases has survived to date, but an intercalary regenerate has formed from the host stump (white colour), distal to which is a pigmented conical structure typical of the distal end of a dA thigh regenerate. Although definite conclusions cannot be based on one case, this result lends support to the idea that all PD levels are represented in the blastema that forms after amputation of a dA thigh. Thus, the truncated appearance of double half-limbs cannot yet be offered as evidence that PD values are serially generated by interactions between transverse values in the amputation plane.

This research was supported by Grant HD-12659 from the National Institutes of Health.

Bryant
,
S. V.
(
1976
).
Regenerative failure of double half limbs in Notophthalmus viridescens
.
Nature
263
,
676
679
.
Bryant
,
S. V.
(
1978
).
Pattern regulation and cell commitment in amphibian limbs
.
In The Clonal Basis of Development
(ed.
S.
Subtelny
&
I. M.
Sussex
), pp.
63
82
.
Academic Press
.
Bryant
,
S. V.
&
Baca
,
B. A.
(
1978
).
Regenerative ability of double half and half upper arms in the newt, Notophthalmus viridescens
.
J. exp. Zool
.
204
,
307
324
.
Butler
,
E. G.
(
1955
).
Regeneration of the urodele limb after reversal of its proximo-distal axis
.
J. Morph
.
96
,
265
282
.
Carlson
,
B. M.
(
1975
).
The effects of rotation and positional change of stump tissues upon morphogenesis of the regenerating axolotl limb
.
Devi Biol
.
47
,
269
291
.
Fallon
,
J. F.
&
Crosby
,
G. M.
(
1977
).
Polarizing zone activity in limb buds of amniotes
.
In Vertebrate Limb and Somite Morphogenesis
(ed.
D. A.
Ede
,
J. R.
Hinchliffe
&
M.
Balls
), pp.
55
69
.
Cambridge University Press
.
French
,
V.
,
Bryant
,
P. J.
&
Bryant
,
S. V.
(
1976
).
Pattern regulation in epimorphic fields
.
Science
193
,
969
981
.
Frost
,
S.
&
Malacinski
,
G. M.
(
1980
).
The developmental genetics of pigment mutants in the Mexican axolotl
.
Devi Gen
.
1
,
271
294
.
Hamburger
,
V.
(
1960
).
A Manual of Experimental Embryology
, p.
196
.
University of Chicago Press
.
Holder
,
N.
&
Tank
,
P. W.
(
1979
).
Morphogenetic interactions occurring between blastemas and stumps after exchanging blastemas between normal and double-half forelimbs in the axolotl, Ambystoma mexicanum
.
Devi Biol
.
68
,
271
279
.
Holder
,
N.
,
Tank
,
P. W.
&
Bryant
,
S. V.
(
1980
).
Regeneration of symmetrical forelimbs in the axolotl, Ambystoma mexicanum
.
Devi Biol
.
74
,
302
314
.
Iten
,
L. E.
&
Bryant
,
S. V.
(
1975
).
The interaction between the blastema and stump in the establishment of the anterior-posterior and proximal-distal organization of the limb regenerate
.
Devi Biol
.
44
,
119
147
.
Krasner
,
G. N.
&
Bryant
,
S. V.
(
1980
).
Distal transformation from double-half forearms in the axolotl, Ambystoma mexicanum
.
Devi Biol
.
74
,
315
325
.
Maden
,
M.
(
1979
).
Regulation and limb regeneration: The effect of partial irradiation
.
J. Embryol. exp. Morph
.
52
,
183
192
.
Maden
,
M.
(
1980
).
Intercalary regeneration in the amphibian limb and the rule of distal transformation
.
J. Embryol. exp. Morph
.
56
,
201
209
.
Pescitelli
,
M. J.
, Jr.
&
Stocum
,
D. L.
(
1980
).
The origin of skeletal structures during intercalary regeneration of larval Ambystoma limbs
.
Dev! Biol
.
79
,
255
275
.
Pescitelli
,
M. J.
, Jr.
&
Stocum
,
D. L.
(
1981
).
Nonsegmental organization of positional information in regenerating Ambystoma limbs
.
Devi Biol
.
82
(in the Press).
Rose
,
S. M.
(
1962
).
Tissue-arc control of regeneration in the amphibian limb
.
In Regeneration
(ed.
D.
Rudnick
), pp.
153
176
.
Ronald Press
.
Shuraleff
,
N. C.
&
Thornton
,
C. S.
(
1967
).
An analysis of distal dominance in the regenerating limb of the axolotl
.
Experientia
23
,
747
752
.
Slack
,
J. M. W.
(
1976
).
Determination of polarity in the amphibian limb
.
Nature (Lond.)
261
,
44
46
.
Slack
,
J. M. W.
(
1977
).
Determination of anteroposterior polarity in the axolotl forelimb by an interaction between limb and flank rudiments
.
J. Embryol. exp. Morph
.
39
,
151
168
.
Slack
,
J. M. W.
(
1980a
).
Regulation and potency in the forelimb rudiment of the axolotl embryo
.
J. Embryol. exp. Morph
.
57
,
203
217
.
Slack
,
J. M. W.
(
1980b
).
A serial threshold theory of regeneration
.
J. theor. Biol
.
82
,
105
140
.
Stocum
,
D. L.
(
1968
).
The urodele limb regeneration blastema: A self-organizing system II. Morphogenesis and differentiation of autografted whole and fractioned blastemas
.
Devi Biol
.
18
,
457
480
.
Stocum
,
D. L.
(
1975
).
Regulation after proximal or distal transposition of limb regeneration blastemas and determination of the proximal boundary of the regenerate
.
Devi Biol
.
45
,
112
136
.
Stocum
,
D. L.
(
1978
).
Regeneration of symmetrical hindlimbs in larval salamanders
.
Science
200
,
790
793
.
Stocum
,
D. L.
(
1979
).
Stages of forelimb regeneration in Ambystoma maculatum
.
J. exp. Zool
.
209
,
395
416
.
Stocum
,
D. L.
(
1980a
).
Intercalary regeneration of symmetrical thighs in the axolotl, Ambystoma mexicanum
.
Devi Biol
.
79
,
276
295
.
Stolcum
,
D. L.
(
1980b
).
Autonomous development of reciprocally exchanged regeneration bastemas of normal forelimbs and symmetrical hindlimbs
.
J. exp. Zool
.
212
,
361
371
.
Summerbell
,
D.
&
Tickle
,
C.
(
1977
).
Pattern formation along the antero-posterior axis of the chick limb bud
.
In Vertebrate Limb and Somite Morphogenesis
(ed.
D. A.
Ede
,
J. R.
Hinchliffe
&
M.
Balls
), pp.
41
53
.
Cambridge University Press
.
Swett
,
F. H.
(
1938
).
Experiments upon the relationships of surrounding areas to polarization of the dorsoventral limb-axis in Ambystoma punctatum (Linn
.).
J. exp. Zool
.
78
,
81
100
.
Tank
,
P. W.
(
1978
).
The failure of double-half forelimbs to undergo distal transformation following amputation in the axolotl, Ambystoma mexicanum
.
J. exp. Zool
.
204
,
325
336
.
Tank
,
P. W.
&
Holder
,
N.
(
1978
).
The effect of healing time on the proximo distal organization of double-half forelimb regenerates in the axolotl, Ambystoma mexicanum
.
Devi Biol
.
66
,
72
85
.
Tank
,
P. W.
&
Holder
,
N.
(
1980
).
The distribution of cells in the upper forelimb of the axolotl, Ambystoma mexicanum
.
J. exp. Zool
.
209
,
435
442
.
Wolpert
,
L.
(
1969
).
Positional information and spatial pattern of cellular differentiation
.
J. theoret. Biol
.
25
,
1
47
.
Wolpert
,
L.
(
1971
).
Positional information and pattern formation
.
Curr. Top. Devi Biol
.
6
,
183
224
.