The emergence of spiraling tracheary element bundles in incompatible grafts

Experimental materials

The seeds of rape (Brassica napus), Nicotiana benthamiana, Arabidopsis thaliana and purslane (Portulaca oleracea) were surface-sterilized in chlorine gas for 2 h. Then they were plated on the sterile MS medium containing 3% sucrose (W/V). Sterile petri dishes containing seeds were vertically placed in the growth room for constant temperature growth (22∼23 °C), and the growth condition was set at long-day photoperiod cycle (16 h of light and 8 h of darkness).

Micrografting

Seedlings germinated after 7-9 days on MS medium were selected for grafting (purslane as rootstock, rape, Nicotiana benthamiana and Arabidopsis as scion, respectively). The grafting process was described previously (Deng et al., 2021). Grafts were grown on moisturized Whatman paper for 3 days, then gently moved to MS medium containing 3% sucrose (w/v) with forceps and continued to grow in a growth room (16 h light /8 h darkness) at 22−23 °C.

Vasculature isolation, scanning electron microscopy and counting of the spiraling TE bundles

Isolation of vasculature was recently described by Liu et al. (2022) and Deng et al. (2021). The leaves and roots of grafts were removed and soaked in 0.04% saponin solution for 30 min, then washed with PEM solution (50 mM PIPES, 5 mM EGTA, 2.5 mM MgSO4, pH 6.9) three times, and then fixed in 4% paraformaldehyde in PEM buffer for 30 min. The fixed samples were washed with PEM solution and then dissected to remove the tissues adjacent to the vascular bundles under a dissecting microscope. The longitudinally dissected vascular samples were treated with the enzyme cocktail solution consisting of 5 mM 2-(N-morpholino) ethanesulfonic acid (MES), 0.5% (w/v) cellulase (EC 3.2.1.4; Sigma-Aldrich, St. Louis, MO, USA), 0.2% (v/v) pectinase (EC 3.2.1.15; Sigma-Aldrich, St. Louis, MO, USA), 0.12 M sucrose, 1 mM CaCl₂(pH 5.5) at 28 °C for 1.5 h. The treated samples were washed in 5% TritonX-100 for 15 min and then in PEM solution for 15 min. Samples were dehydrated step by step (15 min each) with 15%, 30%, 50%, 75% and 100% ethanol. With another two times of absolute ethanol wash, they were dried for 3 h in a −20 °C, using a low-vacuum drier (CHRIST). The dried samples were mounted on the sample holder, and then placed in the ion sputtering apparatus (SC7620 Sputter coater) for conductive gold coating. Examination of the samples was conducted in a VEGA3 TESCAN scanning electron microscope at 20 kv acceleration voltage. In order to count for the number of spiraling bundles, we surveyed the area within a three mm radius centered at the midpoint of the grafting interface. The spiraling bundle was characterized with spiraling tracheary elements plus either a concave hole or convex point, and thus was counted as a spiraling bundle. The Arabidopsis/purslane grafts (n = 67) were surveyed every two days after grafting (DAG), while rape/purslane grafts (n = 35) and Nicotiana/purslane grafts (n = 43) were mainly surveyed after 28 DAG. For each time of collection, graft unions from at least six individual grafts were prepared.

Three interfamilial grafts using purslane as rootstock

In this experiment, we constructed three interfamilial grafts, the Bn/Po heterograft (Figs. 1A and 1B), At/Po heterograft (Figs. 1D and 1E), and Nb/Po heterograft (Figs. 1H and 1I), and four self-grafts, i.e., the Bn (Fig. 1C), At (Fig. 1F), Po (Fig. 1G) and Nb (Fig. 1J) self-graft by micrografting technique. Nearly all the Bn/Po grafts remained quiescent (Fig. 1A), a growth status similar to some grafts of At/Nb combination described previously (Deng et al., 2021), suggesting the two species were highly incompatible. The At/Po grafts, however, grew very quickly at 20 DAG, suggesting that Arabidopsis and purslane was potentially compatible with each other at this stage (Fig. 1D). Usually, a delayed incompatibility was observed around 30 DAG; either the grafts ceased growth, or the grafts developed adventitious roots at the graft union (Table 1). The Nb/Po grafts displayed yellowing of foliage (Fig. 1H, Table 1), suggesting poor development of scion shoot. In addition, the graft unions in both Bn/Po and Nb/Po grafts were enlarged (Figs. 1B and 1I), implying a structural expansion within the unions. In the compatible self-grafts of Bn, At, Po and Nb, the growth of each species was normal (Figs. 1C, 1F, 1G and 1J).

Emergence of spiraling TE bundles or circular TE at the graft union

The scion and the rootstock of the Bn/Po heterografts were easily parted during the preparation for SEM, implying fragile connection between scion and rootstock. The SEM examination showed that spiraling TE bundles extensively existed at the graft union of Bn/Po combination (Fig. 2A), and around 80% of grafts produced spiraling TE bundles at the graft union (Fig. 2B). These spiraling bundles were mainly confined to the upper part of grafting interface, and they strongly rejected the connection with TEs from rootstock. A single circular tracheary element could also be formed through self-fusion as shown in Fig. 2A (shown in red circle), and a repeated process of circling led to the formation of spiraling bundles (Fig. 2A). Apparently, these spiraling bundles lost their TE end, and did not provide a tapered or inclined end wall for overlapping or fusion as normal TE did, therefore could not make a connection with the TEs from rootstock.

In At/Po grafts, the frequency of spiraling TE bundle formation was relatively low (Figs. 2C and 2D), and about 16.7% of grafts (10 out of 60) harbored circular TEs (Fig. 2E). Moreover, the TEs were loosely spiraled (Fig. 2C), thus they usually existed as a single ring and rarely formed a highly spiraling structure (Figs. 2C and 2D). In fact, the At/Po grafts grew well in the first 20 days, agreeing with the less-forming spiraling bundles.

We further examined the spiraling TEs in the scion of Nicotiana benthamiana belonging to solanaceae family, diverging more than 100 million years ago from brassicaceae family according to Ku et al. (2000). Similarly, the enlarged union was consisted of more spiraling TEs (Fig. 2F) which were either convolute concave (Fig. 2G) or convex (Fig. 2H). In an extreme case, these small spirals were fused to form chignon-like structure (Fig. 2I). More than 65% of these grafts contained various numbers of spirals, of whom the majority were two spirals (Fig. 2J). In all compatible self-grafts, there was no spiraling TE bundle detected (Table 1). Taken together, the more compatible At/Po combination has less spiraling TEs than in Bn/Po and Nb/Po combination (Fig. 2K, Table 1), suggesting a close association of graft incompatibility with emergence of spiraling TEs.