Role of HEMIVENATA and the Ubiquitin Pathway in Venation Pattern Formation

Abstract

Elegant work by others has highlighted the importance of auxin transport in venation patterning, an idea substantiated by the severe effects of auxin polar transport inhibitors and by the mutant phenotype and expression patterns associated with the auxin efflux transporter PIN-FORMED1 (PIN1). It is striking, therefore, that little attention has been paid to the venation patterns of mutants insensitive to this hormone, since both auxin transport and perception are crucial components in theoretical models of vascular patterning. Our finding that HEMIVENATA (HVE) is the same gene as CAND1 confirms the role of ubiquitin-mediated auxin perception in vascular patterning and sets the stage for a re-examination of the leaf venation phenotypes of other auxin-resistant mutants and additional components of the ubiquitin pathway.

Recessive hemivenata (hve) alleles confer a pleiotropic phenotype that includes simple leaf venation patterns, with fewer branching points and the absence of most higher-order veins.¹ We previously identified the spontaneous hve-1 allele in the Eiffel-5 wild-type accession. Following a positional approach, we have isolated the gene and found it to be identical to CAND1, also known as ETA2, whose mutations cause mild insensitivity to auxin.^2–4 Loss-of-function hve alleles cause the absence of most tertiary and higher order veins, which is best visualized as a failure to express a provascular reporter gene (ATHB-8-GUS⁵) in the regions where these veins normally differentiate. Consistent with an early function in vascular patterning, the expression of the gene progressively becomes restricted to the vasculature at early stages of leaf development.

CAND1 proteins bind to unneddylated CULLIN1 (CUL1) and regulate the formation of SCF (Skp1-Cul1-F-box protein) ubiquitin ligase complexes.^3,4,6 In agreement with this, we also observed a simple venation pattern associated with the loss of function of AUXIN RESISTANT1 (AXR1), which is necessary for the neddylation of CUL1 in Arabidopsis thaliana.⁷ The similar structural complexity of the venation patterns of hve and axr1 single and double mutants indicates that the post-translational modification of CUL1 is essential for the correct spacing of leaf veins. Unfortunately, the embryonic lethality caused by null alleles of CUL1 makes the study of their adult phenotypes difficult.⁸ Semidominant auxin resistant6 mutations, however, have been shown to be antimorphic alleles of CUL1⁹ and, therefore, the altered venation of axr6 seedlings possibly mimics the effects of loss of CUL1 function. The expected opposite effects of AXR1 and HvE/CAND1 on CUL1 neddylation are in open contradiction with the similarity of the observed loss-of-function phenotypes of axr1 and hve mutants, as also noticed by previous authors.^2,3 If, as has been proposed, functional CAND1 acts as a negative regulator of CUL1 neddylation and hinders its incorporation into SCF complexes, one might predict extra CUL1 neddylation and more SCF complexes to be present in hve mutants. Surprisingly, although neddylation is defective in axr1 mutants, their pleiotropic phenotype includes auxin resistance, simple venation patterns and bushy inflorescences that are comparable to those of hve. We hypothesize that CAND1 may ensure that only neddylated CUL1 incorporates into SCF complexes. In this case, unneddylated CUL1 would interfere with the normal activity of SCF complexes, causing hve mutants to be phenotypically similar to hypomorphic axr1 and antimorphic axr6 mutants.

The scarcity of EMS-induced mutants with normal leaf shape but abnormal vein networks led us to propose that the disruption of venation patterning is masked by redundant genes or pathways, by lethal effects, or by the close coupling of vascular differentiation and other aspects of plant morphogenesis.¹ In agreement with this idea, most venation pattern mutations, including hve and axr1, affect other aspects of plant development pleiotropically. In other cases, the absence of a vascular phenotype can be attributed to functional redundancy of the genes involved. An elegant example of redundancy has recently been reported for members of the YUCCA (YUC) family of flavin monooxygenases, which are involved in auxin biosynthesis.¹⁰ Reduced amounts of auxin in double, triple and quadruple yuc mutants led to simple venation patterns. Like hve mutants, yuc mutants are more drastically affected in higher order veins, showing that similar vascular patterns can arise from altered auxin biosynthesis and perception. Our knowledge of how auxin is perceived has advanced significantly with the identification of the F-box protein TIR1 as the auxin receptor.^11,12 TIR1 assembles into SCF^TIR1 ubiquitin ligase complexes and regulates the ubiquitylation and proteolysis of members of the Aux/IAA family of transcription factors. In remarkable contrast with the lethality of cul1 null alleles, tir1 mutants are only moderately insensitive to exogenously applied auxin¹³ and, to our knowledge, no vascular defects have yet been reported, suggesting extensive redundancy with other closely related members of the same gene family.¹⁴ Although no vascular patterning have been described yet for tir1 single mutants, we would not be surprised if they appear in double or triple? mutant combinations.

One step ahead in the pathway, Aux/IAA proteins physically interact withtranscription factors of the AUxINRESPONSE FACTOR (ARF) family. The BODENLOS/IAA12 protein has been shown to interact with MONOPTEROS (MP/ARF5) and NONPHOTOTROPIC HYPOCOTYL (NPH4/ARF7), two ARF proteins that reportedly function in the patterning of the vasculature.¹⁵ Loss of function of these ARF genes also causes a reduction in the complexity of vascular patterns. Considering that disrupted SCF^TIR1 function (as in axr1 and hve mutants) results in an increased half-life of Aux/IAA proteins, these observations are consistent with Aux/IAA proteins being negative regulators of ARF proteins. These and other mutants with similarly simple vascular networks affect genes that can be classified into at least three different functional categories (auxin biosynthesis, perception and response; Fig. 1). The canalization hypothesis predicts that the formation of discrete vascular strands requires a positive feedback of auxin on its own transport.¹⁶ Interestingly, the polar localization of the auxin transporter PIN-FORMED1 (PIN1) along cell files is among the earliest detectable events during vascular differentiation. The polar intracellular localization of PIN1 has recently been found to depend on the normal function of some Aux/IAA and ARF genes,¹⁷ suggesting that the auxin perception pathway involving TIR1, AxR1, CUL1 and HVE participates in the feedback regulation. Other vascular-specific genes, such as the class III homeodomain-leucine zipper gene ATHB-8, whose expression was previously found to depend on the expression of MP in mutants and over-expressing lines,¹⁸ may not be involved in the establishment of the venation pattern, but instead may be necessary for the differentiation of specific cell types within the veins.

Figure 1.

Venation patterning genes can be classified into functional categories based on their relationship with auxin. Mutants with impaired auxin synthesis, perception and response show similarly reduced vascular patterns. Auxin biosynthesis genes include recently described members of the YUCCA family of flavin monooxygenases. The auxin perception category comprises HEMIVENATA and other components of the neddylation and ubiquitin pathways. The auxin response category includes at least an Aux/IAA protein, BODENLOS, which is targeted by the ubiquitin pathway, and its interactors, which belong to the AUXIN RESPONSE FACTOR family. The localization of the auxin transporter PIN1 has been shown to depend on the function certain Aux/IAA and ARF proteins. The positive feedback loop of auxin on its own transport, proposed by the canalization hypothesis, may depend on similar interactions during the patterning of leaf veins.

Research on venation pattern formation in plant leaves is advancing rapidly, with novel tools and recent reports describing mutants whose auxin metabolism, transport and perception are affected. Further genetic and molecular analyses of those mutants will certainly increase our knowledge in this fascinating field.

Addendum to: Alonso-Peral MM, Candela H, del Pozo JC, Martínez-Laborda A, Ponce MR, Micol JL. The HVE/CAND1 Gene is Required for the Early Patterning of Leaf Venation in Arabidopsis. Development. 2006;133:3755–3766. doi: 10.1242/dev.02554.